{ "source": "https://quantized.uk", "license": "CC BY 4.0", "licenseUrl": "https://creativecommons.org/licenses/by/4.0/", "attribution": "quantized.uk (https://quantized.uk/) — link back when you reuse this data", "termsUrl": "https://quantized.uk/legal/", "scope": "This site's own index and arithmetic. Model weights keep their own licences.", "version": 1, "dataLastUpdated": "2026-10-20", "count": 112, "notes": { "weightsGB": "params × bpw / 8, decimal GB — the weights only. Runtime memory adds KV cache and buffers; use calculatorUrl.", "confidence": "'measured' = run on this site's hardware; 'estimated' = derived from the architecture or a published file size.", "pplLossPercent": "Perplexity increase vs the unquantized weights, where published. Absent means nobody published one.", "arch": "Fields the KV-cache arithmetic reads. attention.fullLayers / windowLayers describe hybrid and sliding-window models; windowTokens is what llama.cpp allocates." }, "models": [ { "id": "llama-3.1-8b", "name": "Llama 3.1 8B Instruct", "family": "Meta Llama 3.1", "paramsB": 8.03, "paramLabel": "8B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q2_K", "bpw": 2.63, "weightsGB": 2.64, "confidence": "measured", "pplLossPercent": 48.5 }, { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.87, "confidence": "measured", "pplLossPercent": 3.2 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 6.58, "confidence": "measured", "pplLossPercent": 0.8 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 8.53, "confidence": "measured", "pplLossPercent": 0.1 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.01, "confidence": "measured", "pplLossPercent": 4.5 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.67, "confidence": "measured", "pplLossPercent": 2.5 } ], "url": "https://quantized.uk/quant-hub/llama-3.1-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.1-8b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-3.1-70b", "name": "Llama 3.1 70B Instruct", "family": "Meta Llama 3.1", "paramsB": 70.6, "paramLabel": "70B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 80, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 42.8, "confidence": "estimated", "pplLossPercent": 2.8 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 50.13, "confidence": "estimated", "pplLossPercent": 1.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 35.3, "confidence": "estimated", "pplLossPercent": 3.9 }, { "format": "EXL2", "level": "3.5bpw", "bpw": 3.5, "weightsGB": 30.89, "confidence": "estimated", "pplLossPercent": 5.2 } ], "url": "https://quantized.uk/quant-hub/llama-3.1-70b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.1-70b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-3.2-3b", "name": "Llama 3.2 3B Instruct", "family": "Meta Llama 3.2", "paramsB": 3.21, "paramLabel": "3B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 28, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 1.95, "confidence": "estimated", "pplLossPercent": 3.5 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 3.41, "confidence": "estimated", "pplLossPercent": 0.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 1.61, "confidence": "estimated", "pplLossPercent": 4.8 } ], "url": "https://quantized.uk/quant-hub/llama-3.2-3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.2-3b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-7b", "name": "Qwen2.5 7B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 7.62, "paramLabel": "7B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 28, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.62, "confidence": "measured", "pplLossPercent": 3 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 6.25, "confidence": "measured", "pplLossPercent": 0.7 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 3.81, "confidence": "measured", "pplLossPercent": 4.2 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.43, "confidence": "measured", "pplLossPercent": 2.2 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-7b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-14b", "name": "Qwen2.5 14B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 14.7, "paramLabel": "14B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 48, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 8.91, "confidence": "estimated", "pplLossPercent": 2.9 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 10.44, "confidence": "estimated", "pplLossPercent": 1.4 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 7.35, "confidence": "estimated", "pplLossPercent": 3.8 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 8.54, "confidence": "estimated", "pplLossPercent": 2.1 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-14b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-14b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-32b", "name": "Qwen2.5 32B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 32.5, "paramLabel": "32B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 64, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 15.72, "confidence": "measured", "pplLossPercent": 7.8 }, { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 19.7, "confidence": "measured", "pplLossPercent": 2.7 }, { "format": "EXL2", "level": "3.5bpw", "bpw": 3.5, "weightsGB": 14.22, "confidence": "measured", "pplLossPercent": 4.8 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-32b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-32b&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-coder-v2-lite", "name": "DeepSeek-Coder-V2-Lite Instruct", "family": "DeepSeek", "paramsB": 15.7, "paramLabel": "16B-A2.4B", "activeParamsB": 2.4, "contextLength": 163840, "categories": [ "code", "instruct" ], "status": "active", "arch": { "layers": 27, "kvHeads": 1, "headDim": 288 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 9.52, "confidence": "estimated", "pplLossPercent": 3.1 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 16.68, "confidence": "estimated", "pplLossPercent": 0.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 7.85, "confidence": "estimated", "pplLossPercent": 4.1 } ], "url": "https://quantized.uk/quant-hub/deepseek-coder-v2-lite/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-coder-v2-lite&quant=Q4_K_M&ctx=4096" }, { "id": "phi-3.5-mini", "name": "Phi-3.5 Mini Instruct", "family": "Microsoft Phi", "paramsB": 3.82, "paramLabel": "3.8B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 32, "kvHeads": 32, "headDim": 96 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 2.32, "confidence": "estimated", "pplLossPercent": 3.8 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 4.06, "confidence": "estimated", "pplLossPercent": 0.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 1.91, "confidence": "estimated", "pplLossPercent": 5.1 } ], "url": "https://quantized.uk/quant-hub/phi-3.5-mini/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=phi-3.5-mini&quant=Q4_K_M&ctx=4096" }, { "id": "mistral-nemo-12b", "name": "Mistral Nemo 12B Instruct", "family": "Mistral AI", "paramsB": 12.2, "paramLabel": "12B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 7.4, "confidence": "estimated", "pplLossPercent": 3.1 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 10, "confidence": "estimated", "pplLossPercent": 0.9 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 6.1, "confidence": "estimated", "pplLossPercent": 4.4 } ], "url": "https://quantized.uk/quant-hub/mistral-nemo-12b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=mistral-nemo-12b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-2-9b", "name": "Gemma 2 9B Instruct", "family": "Google Gemma 2", "paramsB": 9.24, "paramLabel": "9B", "contextLength": 8192, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "gemma-4-e4b", "arch": { "layers": 42, "kvHeads": 8, "headDim": 256, "attention": { "fullLayers": 21, "windowLayers": 21, "windowTokens": 4608 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 5.6, "confidence": "estimated", "pplLossPercent": 3.3 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 9.82, "confidence": "estimated", "pplLossPercent": 0.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.62, "confidence": "estimated", "pplLossPercent": 4.6 } ], "url": "https://quantized.uk/quant-hub/gemma-2-9b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-2-9b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-72b", "name": "Qwen2.5 72B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 72.7, "paramLabel": "72B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 80, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 44.07, "confidence": "estimated", "pplLossPercent": 2.5 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 51.62, "confidence": "estimated", "pplLossPercent": 1.1 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 36.35, "confidence": "estimated", "pplLossPercent": 3.5 }, { "format": "EXL2", "level": "3.5bpw", "bpw": 3.5, "weightsGB": 31.81, "confidence": "estimated", "pplLossPercent": 4.8 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-72b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-72b&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-r1-distill-qwen-14b", "name": "DeepSeek-R1-Distill-Qwen-14B", "family": "DeepSeek", "paramsB": 14.7, "paramLabel": "14B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 48, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 8.91, "confidence": "measured", "pplLossPercent": 2.8 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 8.54, "confidence": "measured", "pplLossPercent": 2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 7.35, "confidence": "measured", "pplLossPercent": 3.6 } ], "url": "https://quantized.uk/quant-hub/deepseek-r1-distill-qwen-14b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-r1-distill-qwen-14b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-3.3-70b", "name": "Llama 3.3 70B Instruct", "family": "Meta Llama 3.3", "paramsB": 70.6, "paramLabel": "70B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 80, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 42.8, "confidence": "estimated", "pplLossPercent": 2.6 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 50.13, "confidence": "estimated", "pplLossPercent": 1 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 35.3, "confidence": "estimated", "pplLossPercent": 3.7 } ], "url": "https://quantized.uk/quant-hub/llama-3.3-70b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.3-70b&quant=Q4_K_M&ctx=4096" }, { "id": "mistral-small-24b", "name": "Mistral Small 24B Instruct", "family": "Mistral AI", "paramsB": 24, "paramLabel": "24B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 14.55, "confidence": "estimated", "pplLossPercent": 2.9 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 12, "confidence": "estimated", "pplLossPercent": 3.8 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 13.95, "confidence": "estimated", "pplLossPercent": 2.2 } ], "url": "https://quantized.uk/quant-hub/mistral-small-24b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=mistral-small-24b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-coder-32b", "name": "Qwen2.5-Coder 32B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 32.5, "paramLabel": "32B", "contextLength": 131072, "categories": [ "code", "instruct" ], "status": "active", "arch": { "layers": 64, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 19.7, "confidence": "estimated", "pplLossPercent": 2.5 }, { "format": "EXL2", "level": "3.5bpw", "bpw": 3.5, "weightsGB": 14.22, "confidence": "estimated", "pplLossPercent": 4.5 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 16.25, "confidence": "estimated", "pplLossPercent": 3.5 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-coder-32b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-coder-32b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-coder-7b", "name": "Qwen2.5-Coder 7B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 7.62, "paramLabel": "7B", "contextLength": 131072, "categories": [ "code", "instruct" ], "status": "active", "arch": { "layers": 28, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.62, "confidence": "estimated", "pplLossPercent": 2.8 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 3.81, "confidence": "estimated", "pplLossPercent": 4 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.43, "confidence": "estimated", "pplLossPercent": 2 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-coder-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-coder-7b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-3b", "name": "Qwen2.5 3B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 3.09, "paramLabel": "3B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 36, "kvHeads": 2, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 1.87, "confidence": "estimated", "pplLossPercent": 3.2 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 3.28, "confidence": "estimated", "pplLossPercent": 0.3 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-3b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-3.2-1b", "name": "Llama 3.2 1B Instruct", "family": "Meta Llama 3.2", "paramsB": 1.24, "paramLabel": "1B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 16, "kvHeads": 8, "headDim": 64 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 0.75, "confidence": "estimated", "pplLossPercent": 4.2 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 1.32, "confidence": "estimated", "pplLossPercent": 0.5 } ], "url": "https://quantized.uk/quant-hub/llama-3.2-1b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.2-1b&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-r1-distill-llama-70b", "name": "DeepSeek-R1-Distill-Llama-70B", "family": "DeepSeek", "paramsB": 70.6, "paramLabel": "70B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 80, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 42.8, "confidence": "estimated", "pplLossPercent": 2.4 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 35.3, "confidence": "estimated", "pplLossPercent": 3.5 } ], "url": "https://quantized.uk/quant-hub/deepseek-r1-distill-llama-70b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-r1-distill-llama-70b&quant=Q4_K_M&ctx=4096" }, { "id": "codestral-22b", "name": "Codestral 22B", "family": "Mistral AI", "paramsB": 22.2, "paramLabel": "22B", "contextLength": 32768, "categories": [ "code", "instruct" ], "status": "active", "arch": { "layers": 56, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 13.46, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 11.1, "confidence": "estimated", "pplLossPercent": 4.2 } ], "url": "https://quantized.uk/quant-hub/codestral-22b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=codestral-22b&quant=Q4_K_M&ctx=4096" }, { "id": "mixtral-8x7b", "name": "Mixtral 8x7B Instruct", "family": "Mistral AI", "paramsB": 46.7, "paramLabel": "47B MoE", "activeParamsB": 12.9, "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-30b-a3b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 28.31, "confidence": "estimated", "pplLossPercent": 2.8 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 23.35, "confidence": "estimated", "pplLossPercent": 3.8 } ], "url": "https://quantized.uk/quant-hub/mixtral-8x7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=mixtral-8x7b&quant=Q4_K_M&ctx=4096" }, { "id": "command-r-35b", "name": "Command R 35B", "family": "Cohere", "paramsB": 35, "paramLabel": "35B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-8-27b", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 21.22, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "GPTQ", "level": "INT4", "bpw": 4, "weightsGB": 17.5, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/command-r-35b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=command-r-35b&quant=Q4_K_M&ctx=4096" }, { "id": "yi-1.5-34b", "name": "Yi 1.5 34B Chat", "family": "01.AI Yi", "paramsB": 34.4, "paramLabel": "34B", "contextLength": 4096, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-32b", "arch": { "layers": 60, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 20.85, "confidence": "estimated", "pplLossPercent": 2.8 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 17.2, "confidence": "estimated", "pplLossPercent": 4 } ], "url": "https://quantized.uk/quant-hub/yi-1.5-34b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=yi-1.5-34b&quant=Q4_K_M&ctx=4096" }, { "id": "solar-10.7b", "name": "Solar 10.7B Instruct", "family": "Upstage", "paramsB": 10.7, "paramLabel": "11B", "contextLength": 4096, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-14b", "arch": { "layers": 48, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 6.49, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 5.35, "confidence": "estimated", "pplLossPercent": 4.2 } ], "url": "https://quantized.uk/quant-hub/solar-10.7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=solar-10.7b&quant=Q4_K_M&ctx=4096" }, { "id": "starcoder2-15b", "name": "StarCoder2 15B", "family": "BigCode", "paramsB": 15, "paramLabel": "15B", "contextLength": 16384, "categories": [ "code", "instruct" ], "status": "superseded", "supersededBy": "qwen2.5-coder-7b", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 9.09, "confidence": "estimated", "pplLossPercent": 3.2 }, { "format": "GPTQ", "level": "INT4", "bpw": 4, "weightsGB": 7.5, "confidence": "estimated", "pplLossPercent": 4.8 } ], "url": "https://quantized.uk/quant-hub/starcoder2-15b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=starcoder2-15b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-3.2-11b-vision", "name": "Llama 3.2 11B Vision Instruct", "family": "Meta Llama 3.2", "paramsB": 11, "paramLabel": "11B", "contextLength": 131072, "categories": [ "multimodal", "instruct" ], "status": "active", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 6.67, "confidence": "estimated", "pplLossPercent": 3.5 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 11.69, "confidence": "estimated", "pplLossPercent": 0.5 } ], "url": "https://quantized.uk/quant-hub/llama-3.2-11b-vision/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.2-11b-vision&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2-vl-7b", "name": "Qwen2-VL 7B Instruct", "family": "Alibaba Qwen2", "paramsB": 8.29, "paramLabel": "7B", "contextLength": 32768, "categories": [ "multimodal", "instruct" ], "status": "superseded", "supersededBy": "qwen3-vl-8b", "arch": { "layers": 28, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 5.03, "confidence": "estimated", "pplLossPercent": 3.8 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.14, "confidence": "estimated", "pplLossPercent": 5 } ], "url": "https://quantized.uk/quant-hub/qwen2-vl-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2-vl-7b&quant=Q4_K_M&ctx=4096" }, { "id": "nous-hermes-3-llama-3.1-8b", "name": "Nous Hermes 3 Llama 3.1 8B", "family": "NousResearch", "paramsB": 8.03, "paramLabel": "8B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.87, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.67, "confidence": "estimated", "pplLossPercent": 2.3 } ], "url": "https://quantized.uk/quant-hub/nous-hermes-3-llama-3.1-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=nous-hermes-3-llama-3.1-8b&quant=Q4_K_M&ctx=4096" }, { "id": "wizardlm-2-7b", "name": "WizardLM-2 7B", "family": "Microsoft / WizardLM", "paramsB": 7.24, "paramLabel": "7B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-8b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.39, "confidence": "estimated", "pplLossPercent": 3.1 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 3.62, "confidence": "estimated", "pplLossPercent": 4.3 } ], "url": "https://quantized.uk/quant-hub/wizardlm-2-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=wizardlm-2-7b&quant=Q4_K_M&ctx=4096" }, { "id": "granite-3.1-8b", "name": "Granite 3.1 8B Instruct", "family": "IBM Granite", "paramsB": 8.2, "paramLabel": "8B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.97, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "GPTQ", "level": "INT4", "bpw": 4, "weightsGB": 4.1, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/granite-3.1-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=granite-3.1-8b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-2-2b", "name": "Gemma 2 2B Instruct", "family": "Google Gemma 2", "paramsB": 2.61, "paramLabel": "2B", "contextLength": 8192, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "gemma-4-e2b", "arch": { "layers": 26, "kvHeads": 4, "headDim": 256, "attention": { "fullLayers": 13, "windowLayers": 13, "windowTokens": 4608 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 1.58, "confidence": "estimated", "pplLossPercent": 3.8 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 2.77, "confidence": "estimated", "pplLossPercent": 0.3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 1.31, "confidence": "estimated", "pplLossPercent": 5 } ], "url": "https://quantized.uk/quant-hub/gemma-2-2b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-2-2b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-2-27b", "name": "Gemma 2 27B Instruct", "family": "Google Gemma 2", "paramsB": 27.2, "paramLabel": "27B", "contextLength": 8192, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "gemma-3-27b-it", "arch": { "layers": 46, "kvHeads": 16, "headDim": 128, "attention": { "fullLayers": 23, "windowLayers": 23, "windowTokens": 4608 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 16.49, "confidence": "estimated", "pplLossPercent": 2.9 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 19.31, "confidence": "estimated", "pplLossPercent": 1.3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 13.6, "confidence": "estimated", "pplLossPercent": 4 } ], "url": "https://quantized.uk/quant-hub/gemma-2-27b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-2-27b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-0.5b", "name": "Qwen2.5 0.5B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 0.49, "paramLabel": "0.5B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 24, "kvHeads": 2, "headDim": 64 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 0.3, "confidence": "estimated", "pplLossPercent": 5.2 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 0.52, "confidence": "estimated", "pplLossPercent": 0.5 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-0.5b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-0.5b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen2.5-1.5b", "name": "Qwen2.5 1.5B Instruct", "family": "Alibaba Qwen2.5", "paramsB": 1.54, "paramLabel": "1.5B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 28, "kvHeads": 2, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 0.93, "confidence": "estimated", "pplLossPercent": 4 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 1.64, "confidence": "estimated", "pplLossPercent": 0.4 } ], "url": "https://quantized.uk/quant-hub/qwen2.5-1.5b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen2.5-1.5b&quant=Q4_K_M&ctx=4096" }, { "id": "phi-3-medium-14b", "name": "Phi-3 Medium 14B Instruct", "family": "Microsoft Phi", "paramsB": 14, "paramLabel": "14B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 40, "kvHeads": 10, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 8.49, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 11.48, "confidence": "estimated", "pplLossPercent": 0.8 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 7, "confidence": "estimated", "pplLossPercent": 4.2 } ], "url": "https://quantized.uk/quant-hub/phi-3-medium-14b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=phi-3-medium-14b&quant=Q4_K_M&ctx=4096" }, { "id": "phi-4-mini", "name": "Phi-4 Mini Instruct", "family": "Microsoft Phi", "paramsB": 3.84, "paramLabel": "3.8B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 2.33, "confidence": "estimated", "pplLossPercent": 3.5 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 4.08, "confidence": "estimated", "pplLossPercent": 0.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 1.92, "confidence": "estimated", "pplLossPercent": 4.8 } ], "url": "https://quantized.uk/quant-hub/phi-4-mini/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=phi-4-mini&quant=Q4_K_M&ctx=4096" }, { "id": "mistral-7b-v0.3", "name": "Mistral 7B Instruct v0.3", "family": "Mistral AI", "paramsB": 7.24, "paramLabel": "7B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "ministral-3-8b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.39, "confidence": "estimated", "pplLossPercent": 3.2 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 5.94, "confidence": "estimated", "pplLossPercent": 0.9 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 3.62, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/mistral-7b-v0.3/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=mistral-7b-v0.3&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-v2-lite", "name": "DeepSeek-V2-Lite Chat", "family": "DeepSeek", "paramsB": 15.7, "paramLabel": "16B-A2.4B", "activeParamsB": 2.4, "contextLength": 163840, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 27, "kvHeads": 1, "headDim": 288 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 9.52, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 7.85, "confidence": "estimated", "pplLossPercent": 4 } ], "url": "https://quantized.uk/quant-hub/deepseek-v2-lite/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-v2-lite&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-r1-distill-qwen-7b", "name": "DeepSeek-R1-Distill-Qwen-7B", "family": "DeepSeek", "paramsB": 7.62, "paramLabel": "7B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 28, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.62, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.43, "confidence": "estimated", "pplLossPercent": 2.2 } ], "url": "https://quantized.uk/quant-hub/deepseek-r1-distill-qwen-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-r1-distill-qwen-7b&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-r1-distill-qwen-32b", "name": "DeepSeek-R1-Distill-Qwen-32B", "family": "DeepSeek", "paramsB": 32.5, "paramLabel": "32B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 64, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 15.72, "confidence": "estimated", "pplLossPercent": 7.2 }, { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 19.7, "confidence": "estimated", "pplLossPercent": 2.6 }, { "format": "EXL2", "level": "3.5bpw", "bpw": 3.5, "weightsGB": 14.22, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/deepseek-r1-distill-qwen-32b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-r1-distill-qwen-32b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-3.2-90b-vision", "name": "Llama 3.2 90B Vision Instruct", "family": "Meta Llama 3.2", "paramsB": 88.6, "paramLabel": "90B", "contextLength": 131072, "categories": [ "multimodal", "instruct" ], "status": "active", "arch": { "layers": 80, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 53.71, "confidence": "estimated", "pplLossPercent": 2.8 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 42.86, "confidence": "estimated", "pplLossPercent": 6.5 } ], "url": "https://quantized.uk/quant-hub/llama-3.2-90b-vision/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.2-90b-vision&quant=Q4_K_M&ctx=4096" }, { "id": "olmo-2-7b", "name": "OLMo 2 7B Instruct", "family": "Allen AI OLMo", "paramsB": 7.3, "paramLabel": "7B", "contextLength": 4096, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "olmo-3-7b", "arch": { "layers": 32, "kvHeads": 32, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.43, "confidence": "estimated", "pplLossPercent": 3.4 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 7.76, "confidence": "estimated", "pplLossPercent": 0.3 } ], "url": "https://quantized.uk/quant-hub/olmo-2-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=olmo-2-7b&quant=Q4_K_M&ctx=4096" }, { "id": "internlm2-7b", "name": "InternLM2 7B Chat", "family": "Shanghai AI Lab", "paramsB": 7.74, "paramLabel": "7B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-8b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.69, "confidence": "estimated", "pplLossPercent": 3.1 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 3.87, "confidence": "estimated", "pplLossPercent": 4.3 } ], "url": "https://quantized.uk/quant-hub/internlm2-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=internlm2-7b&quant=Q4_K_M&ctx=4096" }, { "id": "internlm2-20b", "name": "InternLM2 20B Chat", "family": "Shanghai AI Lab", "paramsB": 19.9, "paramLabel": "20B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "gpt-oss-20b", "arch": { "layers": 48, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 12.06, "confidence": "estimated", "pplLossPercent": 2.9 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 14.13, "confidence": "estimated", "pplLossPercent": 1.4 } ], "url": "https://quantized.uk/quant-hub/internlm2-20b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=internlm2-20b&quant=Q4_K_M&ctx=4096" }, { "id": "aya-23-8b", "name": "Aya 23 8B", "family": "Cohere For AI", "paramsB": 8.03, "paramLabel": "8B", "contextLength": 8192, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-8b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.87, "confidence": "estimated", "pplLossPercent": 3.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.01, "confidence": "estimated", "pplLossPercent": 4.6 } ], "url": "https://quantized.uk/quant-hub/aya-23-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=aya-23-8b&quant=Q4_K_M&ctx=4096" }, { "id": "openchat-3.6-8b", "name": "OpenChat 3.6 8B", "family": "OpenChat", "paramsB": 8.03, "paramLabel": "8B", "contextLength": 8192, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-8b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.87, "confidence": "estimated", "pplLossPercent": 3.1 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.67, "confidence": "estimated", "pplLossPercent": 2.4 } ], "url": "https://quantized.uk/quant-hub/openchat-3.6-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=openchat-3.6-8b&quant=Q4_K_M&ctx=4096" }, { "id": "zephyr-7b-beta", "name": "Zephyr 7B Beta", "family": "HuggingFaceH4", "paramsB": 7.24, "paramLabel": "7B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "qwen3-8b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.39, "confidence": "estimated", "pplLossPercent": 3.3 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 5.94, "confidence": "estimated", "pplLossPercent": 1 } ], "url": "https://quantized.uk/quant-hub/zephyr-7b-beta/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=zephyr-7b-beta&quant=Q4_K_M&ctx=4096" }, { "id": "stablelm-2-12b", "name": "Stable LM 2 12B Chat", "family": "Stability AI", "paramsB": 12.1, "paramLabel": "12B", "contextLength": 4096, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "falcon-3-10b", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 7.34, "confidence": "estimated", "pplLossPercent": 3.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 6.05, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/stablelm-2-12b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=stablelm-2-12b&quant=Q4_K_M&ctx=4096" }, { "id": "falcon-3-10b", "name": "Falcon 3 10B Instruct", "family": "TII UAE", "paramsB": 10.4, "paramLabel": "10B", "contextLength": 32768, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 6.31, "confidence": "estimated", "pplLossPercent": 3.1 }, { "format": "GPTQ", "level": "INT4", "bpw": 4, "weightsGB": 5.2, "confidence": "estimated", "pplLossPercent": 4.8 } ], "url": "https://quantized.uk/quant-hub/falcon-3-10b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=falcon-3-10b&quant=Q4_K_M&ctx=4096" }, { "id": "jamba-1.5-mini", "name": "Jamba 1.5 Mini", "family": "AI21 Labs", "paramsB": 52, "paramLabel": "52B-A12B", "activeParamsB": 12, "contextLength": 262144, "categories": [ "general", "instruct" ], "status": "superseded", "supersededBy": "ministral-3-8b", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128, "attention": { "fullLayers": 4 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 31.53, "confidence": "estimated", "pplLossPercent": 3.4 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 26, "confidence": "estimated", "pplLossPercent": 4.8 } ], "url": "https://quantized.uk/quant-hub/jamba-1.5-mini/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=jamba-1.5-mini&quant=Q4_K_M&ctx=4096" }, { "id": "dbrx-instruct", "name": "DBRX Instruct", "family": "Databricks", "paramsB": 132, "paramLabel": "132B MoE", "activeParamsB": 36, "contextLength": 32768, "categories": [ "general", "instruct", "code" ], "status": "superseded", "supersededBy": "glm-4.5-air", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 80.02, "confidence": "estimated", "pplLossPercent": 2.5 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 63.86, "confidence": "estimated", "pplLossPercent": 5.8 } ], "url": "https://quantized.uk/quant-hub/dbrx-instruct/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=dbrx-instruct&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-8b", "name": "Qwen3 8B Instruct", "family": "Alibaba Qwen3", "paramsB": 8.19, "paramLabel": "8B", "contextLength": 40960, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 36, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.97, "confidence": "measured", "pplLossPercent": 2.8 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 6.72, "confidence": "measured", "pplLossPercent": 0.6 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.1, "confidence": "measured", "pplLossPercent": 3.8 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.76, "confidence": "measured", "pplLossPercent": 2 } ], "url": "https://quantized.uk/quant-hub/qwen3-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-8b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-14b", "name": "Qwen3 14B Instruct", "family": "Alibaba Qwen3", "paramsB": 14.77, "paramLabel": "14B", "contextLength": 40960, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 8.95, "confidence": "measured", "pplLossPercent": 2.6 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 10.49, "confidence": "measured", "pplLossPercent": 1.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 7.39, "confidence": "measured", "pplLossPercent": 3.5 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 8.59, "confidence": "measured", "pplLossPercent": 1.8 } ], "url": "https://quantized.uk/quant-hub/qwen3-14b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-14b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-3-4b-it", "name": "Gemma 3 4B IT", "family": "Google Gemma 3", "paramsB": 4.3, "paramLabel": "4B", "contextLength": 131072, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "arch": { "layers": 34, "kvHeads": 4, "headDim": 256, "attention": { "fullLayers": 5, "windowLayers": 29, "windowTokens": 1536 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 2.61, "confidence": "estimated", "pplLossPercent": 3.2 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 4.57, "confidence": "estimated", "pplLossPercent": 0.2 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 2.15, "confidence": "estimated", "pplLossPercent": 4.2 } ], "url": "https://quantized.uk/quant-hub/gemma-3-4b-it/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-3-4b-it&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-3-12b-it", "name": "Gemma 3 12B IT", "family": "Google Gemma 3", "paramsB": 12.2, "paramLabel": "12B", "contextLength": 131072, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "arch": { "layers": 48, "kvHeads": 8, "headDim": 256, "attention": { "fullLayers": 8, "windowLayers": 40, "windowTokens": 1536 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 7.4, "confidence": "estimated", "pplLossPercent": 2.9 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 8.66, "confidence": "estimated", "pplLossPercent": 1.3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 6.1, "confidence": "estimated", "pplLossPercent": 3.9 } ], "url": "https://quantized.uk/quant-hub/gemma-3-12b-it/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-3-12b-it&quant=Q4_K_M&ctx=4096" }, { "id": "llama-4-scout-17b", "name": "Llama 4 Scout 17B (16E)", "family": "Meta Llama 4", "paramsB": 109, "paramLabel": "109B MoE", "activeParamsB": 17, "contextLength": 10485760, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "arch": { "layers": 48, "kvHeads": 8, "headDim": 128, "attention": { "fullLayers": 12, "windowLayers": 36, "windowTokens": 8704 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 66.08, "confidence": "estimated", "pplLossPercent": 2.4 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 52.73, "confidence": "estimated", "pplLossPercent": 4.8 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 54.5, "confidence": "estimated", "pplLossPercent": 3.2 } ], "url": "https://quantized.uk/quant-hub/llama-4-scout-17b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-4-scout-17b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-4-maverick-17b", "name": "Llama 4 Maverick 17B (128E)", "family": "Meta Llama 4", "paramsB": 400, "paramLabel": "400B MoE", "activeParamsB": 17, "contextLength": 1048576, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "arch": { "layers": 48, "kvHeads": 8, "headDim": 128, "attention": { "fullLayers": 12, "windowLayers": 36, "windowTokens": 8704 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 242.5, "confidence": "estimated", "pplLossPercent": 2.2 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 193.5, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/llama-4-maverick-17b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-4-maverick-17b&quant=Q4_K_M&ctx=4096" }, { "id": "llama-3.1-405b", "name": "Llama 3.1 405B Instruct", "family": "Meta Llama 3.1", "paramsB": 405, "paramLabel": "405B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "arch": { "layers": 126, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 245.53, "confidence": "estimated", "pplLossPercent": 2.3 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 195.92, "confidence": "estimated", "pplLossPercent": 5 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 202.5, "confidence": "estimated", "pplLossPercent": 3.5 } ], "url": "https://quantized.uk/quant-hub/llama-3.1-405b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=llama-3.1-405b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-32b", "name": "Qwen3 32B Instruct", "family": "Alibaba Qwen3", "paramsB": 32.76, "paramLabel": "32B", "contextLength": 40960, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 64, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 15.85, "confidence": "measured", "pplLossPercent": 7.2 }, { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 19.86, "confidence": "measured", "pplLossPercent": 2.5 }, { "format": "EXL2", "level": "3.5bpw", "bpw": 3.5, "weightsGB": 14.33, "confidence": "measured", "pplLossPercent": 4.5 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 16.38, "confidence": "measured", "pplLossPercent": 3.6 } ], "url": "https://quantized.uk/quant-hub/qwen3-32b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-32b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-30b-a3b", "name": "Qwen3 30B-A3B Instruct", "family": "Alibaba Qwen3", "paramsB": 30.5, "paramLabel": "30B-A3B", "activeParamsB": 3, "contextLength": 40960, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 48, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 18.49, "confidence": "measured", "pplLossPercent": 2.4 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 21.65, "confidence": "measured", "pplLossPercent": 1.1 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 15.25, "confidence": "measured", "pplLossPercent": 3.4 } ], "url": "https://quantized.uk/quant-hub/qwen3-30b-a3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-30b-a3b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-235b-a22b", "name": "Qwen3 235B-A22B Instruct", "family": "Alibaba Qwen3", "paramsB": 235, "paramLabel": "235B-A22B", "activeParamsB": 22, "contextLength": 40960, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 94, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 142.47, "confidence": "estimated", "pplLossPercent": 2.2 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 113.68, "confidence": "estimated", "pplLossPercent": 4.8 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 117.5, "confidence": "estimated", "pplLossPercent": 3.2 } ], "url": "https://quantized.uk/quant-hub/qwen3-235b-a22b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-235b-a22b&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-v3", "name": "DeepSeek-V3", "family": "DeepSeek", "paramsB": 671, "paramLabel": "671B MoE", "activeParamsB": 37, "contextLength": 163840, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 61, "kvHeads": 1, "headDim": 288 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 406.79, "confidence": "estimated", "pplLossPercent": 2 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 324.6, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/deepseek-v3/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-v3&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-r1", "name": "DeepSeek-R1", "family": "DeepSeek", "paramsB": 671, "paramLabel": "671B MoE", "activeParamsB": 37, "contextLength": 163840, "categories": [ "general", "instruct", "code" ], "status": "active", "arch": { "layers": 61, "kvHeads": 1, "headDim": 288 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 406.79, "confidence": "estimated", "pplLossPercent": 1.8 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 324.6, "confidence": "estimated", "pplLossPercent": 4.2 } ], "url": "https://quantized.uk/quant-hub/deepseek-r1/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-r1&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-4b", "name": "Qwen3 4B Instruct", "family": "Alibaba Qwen3", "paramsB": 4, "paramLabel": "4B", "contextLength": 32768, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-06-26", "arch": { "layers": 36, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 2.42, "confidence": "measured", "pplLossPercent": 2.9 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 3.28, "confidence": "measured", "pplLossPercent": 0.7 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 2, "confidence": "measured", "pplLossPercent": 3.8 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 2.33, "confidence": "measured", "pplLossPercent": 2.1 } ], "url": "https://quantized.uk/quant-hub/qwen3-4b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-4b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-coder-30b-a3b", "name": "Qwen3-Coder 30B-A3B Instruct", "family": "Alibaba Qwen3", "paramsB": 30.5, "paramLabel": "30B-A3B", "activeParamsB": 3, "contextLength": 262144, "categories": [ "code", "instruct" ], "status": "active", "addedToIndex": "2026-06-26", "arch": { "layers": 48, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 18.49, "confidence": "measured", "pplLossPercent": 2.3 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 21.65, "confidence": "measured", "pplLossPercent": 1 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 15.25, "confidence": "measured", "pplLossPercent": 3.2 } ], "url": "https://quantized.uk/quant-hub/qwen3-coder-30b-a3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-coder-30b-a3b&quant=Q4_K_M&ctx=4096" }, { "id": "mistral-large-3", "name": "Mistral Large 3 675B Instruct", "family": "Mistral AI", "paramsB": 675, "paramLabel": "675B MoE", "activeParamsB": 41, "contextLength": 262144, "categories": [ "general", "instruct", "multimodal", "code" ], "status": "active", "addedToIndex": "2026-06-26", "arch": { "layers": 64, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 409.22, "confidence": "estimated", "pplLossPercent": 2.1 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 326.53, "confidence": "estimated", "pplLossPercent": 4.5 } ], "url": "https://quantized.uk/quant-hub/mistral-large-3/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=mistral-large-3&quant=Q4_K_M&ctx=4096" }, { "id": "glm-4-9b-chat", "name": "GLM-4-9B-Chat", "family": "Zhipu GLM-4", "paramsB": 9, "paramLabel": "9B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-06-26", "arch": { "layers": 40, "kvHeads": 2, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 5.46, "confidence": "estimated", "pplLossPercent": 3 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 9.56, "confidence": "estimated", "pplLossPercent": 0.4 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.5, "confidence": "estimated", "pplLossPercent": 4.2 } ], "url": "https://quantized.uk/quant-hub/glm-4-9b-chat/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=glm-4-9b-chat&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-3-27b-it", "name": "Gemma 3 27B IT", "family": "Google Gemma 3", "paramsB": 27.4, "paramLabel": "27B", "contextLength": 131072, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "addedToIndex": "2026-07-22", "arch": { "layers": 62, "kvHeads": 16, "headDim": 128, "attention": { "fullLayers": 10, "windowLayers": 52, "windowTokens": 1536 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 16.61, "confidence": "community", "pplLossPercent": 2.8 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 13.25, "confidence": "estimated", "pplLossPercent": 5.5 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 13.7, "confidence": "estimated", "pplLossPercent": 3.8 } ], "url": "https://quantized.uk/quant-hub/gemma-3-27b-it/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-3-27b-it&quant=Q4_K_M&ctx=4096" }, { "id": "deepseek-r1-distill-llama-8b", "name": "DeepSeek-R1-Distill-Llama-8B", "family": "DeepSeek", "paramsB": 8.03, "paramLabel": "8B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-07-22", "arch": { "layers": 32, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.87, "confidence": "community", "pplLossPercent": 2.6 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 5.7, "confidence": "estimated", "pplLossPercent": 1.2 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 4.67, "confidence": "estimated", "pplLossPercent": 1.9 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.01, "confidence": "estimated", "pplLossPercent": 3.5 } ], "url": "https://quantized.uk/quant-hub/deepseek-r1-distill-llama-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=deepseek-r1-distill-llama-8b&quant=Q4_K_M&ctx=4096" }, { "id": "phi-4", "name": "Phi-4 14B", "family": "Microsoft Phi", "paramsB": 14.7, "paramLabel": "14B", "contextLength": 16384, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-07-22", "arch": { "layers": 40, "kvHeads": 10, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 8.91, "confidence": "community", "pplLossPercent": 2.7 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 10.44, "confidence": "estimated", "pplLossPercent": 1.3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 7.35, "confidence": "estimated", "pplLossPercent": 3.6 } ], "url": "https://quantized.uk/quant-hub/phi-4/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=phi-4&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-1.7b", "name": "Qwen3 1.7B Instruct", "family": "Alibaba Qwen3", "paramsB": 1.72, "paramLabel": "1.7B", "contextLength": 32768, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-07-22", "arch": { "layers": 28, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 1.04, "confidence": "community", "pplLossPercent": 3 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 1.83, "confidence": "estimated", "pplLossPercent": 0.4 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 0.86, "confidence": "estimated", "pplLossPercent": 4 } ], "url": "https://quantized.uk/quant-hub/qwen3-1.7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-1.7b&quant=Q4_K_M&ctx=4096" }, { "id": "gpt-oss-20b", "name": "GPT-OSS 20B", "family": "OpenAI GPT-OSS", "paramsB": 20.9, "paramLabel": "21B MoE", "activeParamsB": 3.6, "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-08-07", "arch": { "layers": 24, "kvHeads": 8, "headDim": 64, "attention": { "fullLayers": 12, "windowLayers": 12, "windowTokens": 768 } }, "referenceLevel": "MXFP4", "quants": [ { "format": "GGUF", "level": "MXFP4", "bpw": 4.63, "weightsGB": 12.1, "confidence": "community", "pplLossPercent": 0 }, { "format": "GGUF", "level": "Q8_0", "bpw": 5.1, "weightsGB": 13.32, "confidence": "estimated", "pplLossPercent": 0 }, { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.1, "weightsGB": 10.71, "confidence": "estimated", "pplLossPercent": 1.4 } ], "url": "https://quantized.uk/quant-hub/gpt-oss-20b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gpt-oss-20b&quant=MXFP4&ctx=4096" }, { "id": "gpt-oss-120b", "name": "GPT-OSS 120B", "family": "OpenAI GPT-OSS", "paramsB": 116.8, "paramLabel": "117B MoE", "activeParamsB": 5.1, "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-08-07", "arch": { "layers": 36, "kvHeads": 8, "headDim": 64, "attention": { "fullLayers": 18, "windowLayers": 18, "windowTokens": 768 } }, "referenceLevel": "MXFP4", "quants": [ { "format": "GGUF", "level": "MXFP4", "bpw": 4.34, "weightsGB": 63.36, "confidence": "community", "pplLossPercent": 0 }, { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.1, "weightsGB": 59.86, "confidence": "estimated", "pplLossPercent": 1.6 } ], "url": "https://quantized.uk/quant-hub/gpt-oss-120b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gpt-oss-120b&quant=MXFP4&ctx=4096" }, { "id": "glm-4.5-air", "name": "GLM-4.5-Air", "family": "Zhipu GLM-4.5", "paramsB": 106, "paramLabel": "106B MoE", "activeParamsB": 12, "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-08-07", "arch": { "layers": 46, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 64.26, "confidence": "community", "pplLossPercent": 2.4 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 51.28, "confidence": "estimated", "pplLossPercent": 5.1 }, { "format": "GGUF", "level": "Q2_K", "bpw": 2.63, "weightsGB": 34.85, "confidence": "estimated", "pplLossPercent": 12 } ], "url": "https://quantized.uk/quant-hub/glm-4.5-air/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=glm-4.5-air&quant=Q4_K_M&ctx=4096" }, { "id": "devstral-small-2507", "name": "Devstral Small 1.1 24B", "family": "Mistral Devstral", "paramsB": 23.6, "paramLabel": "24B", "contextLength": 131072, "categories": [ "code", "instruct" ], "status": "active", "addedToIndex": "2026-08-07", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 14.31, "confidence": "community", "pplLossPercent": 2.9 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 16.76, "confidence": "estimated", "pplLossPercent": 1.4 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 19.35, "confidence": "estimated", "pplLossPercent": 0.6 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 11.8, "confidence": "estimated", "pplLossPercent": 4 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 13.72, "confidence": "estimated", "pplLossPercent": 2.5 } ], "url": "https://quantized.uk/quant-hub/devstral-small-2507/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=devstral-small-2507&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-vl-8b", "name": "Qwen3-VL 8B Instruct", "family": "Alibaba Qwen3-VL", "paramsB": 8.8, "paramLabel": "8B", "contextLength": 40960, "categories": [ "multimodal", "instruct", "general" ], "status": "active", "addedToIndex": "2026-08-08", "arch": { "layers": 36, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 5.34, "confidence": "community", "pplLossPercent": 3 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 6.25, "confidence": "estimated", "pplLossPercent": 1.5 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 9.35, "confidence": "estimated", "pplLossPercent": 0.3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 4.4, "confidence": "estimated", "pplLossPercent": 4.2 } ], "url": "https://quantized.uk/quant-hub/qwen3-vl-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-vl-8b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-vl-30b-a3b", "name": "Qwen3-VL 30B-A3B Instruct", "family": "Alibaba Qwen3-VL", "paramsB": 30.5, "paramLabel": "30B-A3B", "activeParamsB": 3, "contextLength": 40960, "categories": [ "multimodal", "instruct", "general" ], "status": "active", "addedToIndex": "2026-08-08", "arch": { "layers": 48, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 18.49, "confidence": "community", "pplLossPercent": 2.6 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 14.75, "confidence": "estimated", "pplLossPercent": 5.4 }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 32.41, "confidence": "estimated", "pplLossPercent": 0.3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 15.25, "confidence": "estimated", "pplLossPercent": 3.7 } ], "url": "https://quantized.uk/quant-hub/qwen3-vl-30b-a3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-vl-30b-a3b&quant=Q4_K_M&ctx=4096" }, { "id": "magistral-small-2509", "name": "Magistral Small 1.2 24B", "family": "Mistral Magistral", "paramsB": 23.6, "paramLabel": "24B", "contextLength": 131072, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "addedToIndex": "2026-08-08", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 14.31, "confidence": "community", "pplLossPercent": 2.9 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 16.76, "confidence": "estimated", "pplLossPercent": 1.4 }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 19.35, "confidence": "estimated", "pplLossPercent": 0.6 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 11.8, "confidence": "estimated", "pplLossPercent": 4 }, { "format": "EXL2", "level": "4.65bpw", "bpw": 4.65, "weightsGB": 13.72, "confidence": "estimated", "pplLossPercent": 2.5 } ], "url": "https://quantized.uk/quant-hub/magistral-small-2509/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=magistral-small-2509&quant=Q4_K_M&ctx=4096" }, { "id": "seed-oss-36b", "name": "Seed-OSS 36B Instruct", "family": "ByteDance Seed", "paramsB": 36, "paramLabel": "36B", "contextLength": 524288, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-08-08", "arch": { "layers": 64, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 21.83, "confidence": "community", "pplLossPercent": 2.7 }, { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.87, "weightsGB": 17.42, "confidence": "estimated", "pplLossPercent": 5.6 }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 25.56, "confidence": "estimated", "pplLossPercent": 1.3 }, { "format": "AWQ", "level": "INT4", "bpw": 4, "weightsGB": 18, "confidence": "estimated", "pplLossPercent": 3.8 } ], "url": "https://quantized.uk/quant-hub/seed-oss-36b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=seed-oss-36b&quant=Q4_K_M&ctx=4096" }, { "id": "ministral-3-8b", "name": "Ministral 3 8B Instruct", "family": "Mistral Ministral 3", "paramsB": 8.49, "paramLabel": "8B", "contextLength": 262144, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-09-11", "arch": { "layers": 34, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.9, "weightsGB": 5.2, "confidence": "community" }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.71, "weightsGB": 6.06, "confidence": "community" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.51, "weightsGB": 9.03, "confidence": "community" } ], "url": "https://quantized.uk/quant-hub/ministral-3-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=ministral-3-8b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-8-27b", "name": "Qwen3.8 27B", "family": "Alibaba Qwen3.8", "paramsB": 27.78, "paramLabel": "27B", "contextLength": 262144, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-09-11", "arch": { "layers": 64, "kvHeads": 4, "headDim": 256, "attention": { "fullLayers": 16 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.74, "weightsGB": 16.46, "confidence": "community" } ], "url": "https://quantized.uk/quant-hub/qwen3-8-27b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-8-27b&quant=Q4_K_M&ctx=4096" }, { "id": "glm-4.7-flash", "name": "GLM-4.7-Flash", "family": "Zhipu GLM-4.7", "paramsB": 29.94, "paramLabel": "30B-A3B", "activeParamsB": 3, "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-09-30", "arch": { "layers": 47, "kvHeads": 1, "headDim": 288 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 18.15, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 31.81, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/glm-4.7-flash/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=glm-4.7-flash&quant=Q4_K_M&ctx=4096" }, { "id": "kimi-linear-48b-a3b", "name": "Kimi Linear 48B-A3B Instruct", "family": "Moonshot Kimi Linear", "paramsB": 48, "paramLabel": "48B-A3B", "activeParamsB": 3, "contextLength": 1048576, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-09-30", "arch": { "layers": 27, "kvHeads": 1, "headDim": 288, "attention": { "fullLayers": 7 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 29.1, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 51, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/kimi-linear-48b-a3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=kimi-linear-48b-a3b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-4-26b-a4b", "name": "Gemma 4 26B-A4B IT", "family": "Google Gemma 4", "paramsB": 25.23, "paramLabel": "26B-A4B", "activeParamsB": 4, "contextLength": 262144, "categories": [ "general", "instruct", "multimodal", "code" ], "status": "active", "addedToIndex": "2026-10-01", "arch": { "layers": 30, "kvHeads": 8, "headDim": 256, "attention": { "fullLayers": 5, "windowLayers": 25, "windowTokens": 1536, "fullKvHeads": 2, "fullHeadDim": 512 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 15.3, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 26.81, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/gemma-4-26b-a4b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-4-26b-a4b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-4-31b", "name": "Gemma 4 31B IT", "family": "Google Gemma 4", "paramsB": 30.7, "paramLabel": "31B", "contextLength": 262144, "categories": [ "general", "instruct", "multimodal", "code" ], "status": "active", "addedToIndex": "2026-10-01", "arch": { "layers": 60, "kvHeads": 16, "headDim": 256, "attention": { "fullLayers": 10, "windowLayers": 50, "windowTokens": 1536, "fullKvHeads": 4, "fullHeadDim": 512 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 18.61, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 32.62, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/gemma-4-31b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-4-31b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-4-e4b", "name": "Gemma 4 E4B IT", "family": "Google Gemma 4", "paramsB": 7.52, "paramLabel": "E4B", "contextLength": 131072, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "addedToIndex": "2026-10-02", "arch": { "layers": 42, "kvHeads": 2, "headDim": 256, "attention": { "fullLayers": 4, "windowLayers": 20, "windowTokens": 1024, "fullKvHeads": 2, "fullHeadDim": 512 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.56, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 7.99, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/gemma-4-e4b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-4-e4b&quant=Q4_K_M&ctx=4096" }, { "id": "gemma-4-e2b", "name": "Gemma 4 E2B IT", "family": "Google Gemma 4", "paramsB": 4.65, "paramLabel": "E2B", "contextLength": 131072, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "addedToIndex": "2026-10-02", "arch": { "layers": 35, "kvHeads": 1, "headDim": 256, "attention": { "fullLayers": 3, "windowLayers": 12, "windowTokens": 1024, "fullKvHeads": 1, "fullHeadDim": 512 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 2.82, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 4.94, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/gemma-4-e2b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-4-e2b&quant=Q4_K_M&ctx=4096" }, { "id": "smollm3-3b", "name": "SmolLM3 3B", "family": "Hugging Face SmolLM", "paramsB": 3.08, "paramLabel": "3B", "contextLength": 65536, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-10-03", "arch": { "layers": 36, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 1.87, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 3.27, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/smollm3-3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=smollm3-3b&quant=Q4_K_M&ctx=4096" }, { "id": "ernie-4.5-21b-a3b", "name": "ERNIE 4.5 21B-A3B", "family": "Baidu ERNIE 4.5", "paramsB": 21.8, "paramLabel": "21B-A3B", "activeParamsB": 3, "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-10-03", "arch": { "layers": 28, "kvHeads": 4, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 13.22, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 23.16, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/ernie-4.5-21b-a3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=ernie-4.5-21b-a3b&quant=Q4_K_M&ctx=4096" }, { "id": "olmo-3-7b", "name": "OLMo 3 7B Instruct", "family": "Allen AI OLMo", "paramsB": 7.3, "paramLabel": "7B", "contextLength": 65536, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-10-05", "arch": { "layers": 32, "kvHeads": 32, "headDim": 128, "attention": { "fullLayers": 8, "windowLayers": 24, "windowTokens": 4608 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 4.43, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 7.76, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/olmo-3-7b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=olmo-3-7b&quant=Q4_K_M&ctx=4096" }, { "id": "muse-glimmer-30b", "name": "Muse Glimmer 30B", "family": "Meta Muse", "paramsB": 27.85, "paramLabel": "30B", "contextLength": 131072, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-09", "arch": { "layers": 52, "kvHeads": 2, "headDim": 128, "attention": { "fullLayers": 13, "windowLayers": 39, "windowTokens": 2560 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.81, "weightsGB": 16.74, "confidence": "estimated" }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.78, "weightsGB": 20.12, "confidence": "estimated" }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.72, "weightsGB": 23.39, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 29.59, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/muse-glimmer-30b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=muse-glimmer-30b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-8-flash-next", "name": "Qwen3.8-Flash-Next", "family": "Qwen3.8", "paramsB": 125, "paramLabel": "125B-A6B", "activeParamsB": 6, "contextLength": 262144, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-10-10", "arch": { "layers": 48, "kvHeads": 2, "headDim": 256, "attention": { "fullLayers": 12, "windowLayers": 0, "fullHeadDim": 320 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 75.78, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 132.81, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/qwen3-8-flash-next/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-8-flash-next&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-6-35b-a3b", "name": "Qwen3.6 35B-A3B", "family": "Alibaba Qwen3.6", "paramsB": 35.59, "paramLabel": "35B-A3B", "activeParamsB": 3, "contextLength": 262144, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-10", "arch": { "layers": 40, "kvHeads": 2, "headDim": 256, "attention": { "fullLayers": 10 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 5.01, "weightsGB": 22.29, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 37.81, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/qwen3-6-35b-a3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-6-35b-a3b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-6-27b", "name": "Qwen3.6 27B", "family": "Alibaba Qwen3.6", "paramsB": 27.41, "paramLabel": "27B", "contextLength": 262144, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-10", "arch": { "layers": 64, "kvHeads": 4, "headDim": 256, "attention": { "fullLayers": 16 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 5.25, "weightsGB": 17.99, "confidence": "estimated" }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 6.12, "weightsGB": 20.97, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 29.12, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/qwen3-6-27b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-6-27b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-5-9b", "name": "Qwen3.5 9B", "family": "Alibaba Qwen3.5", "paramsB": 9.22, "paramLabel": "9B", "contextLength": 262144, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-10", "arch": { "layers": 32, "kvHeads": 4, "headDim": 256, "attention": { "fullLayers": 8 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 5.35, "weightsGB": 6.17, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 9.8, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/qwen3-5-9b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-5-9b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-5-4b", "name": "Qwen3.5 4B", "family": "Alibaba Qwen3.5", "paramsB": 4.35, "paramLabel": "4B", "contextLength": 262144, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "addedToIndex": "2026-10-11", "arch": { "layers": 32, "kvHeads": 4, "headDim": 256, "attention": { "fullLayers": 8 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 5.54, "weightsGB": 3.01, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 4.62, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/qwen3-5-4b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-5-4b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-5-2b", "name": "Qwen3.5 2B", "family": "Alibaba Qwen3.5", "paramsB": 1.96, "paramLabel": "2B", "contextLength": 262144, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "addedToIndex": "2026-10-11", "arch": { "layers": 24, "kvHeads": 2, "headDim": 256, "attention": { "fullLayers": 6 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 5.71, "weightsGB": 1.4, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.49, "weightsGB": 2.08, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/qwen3-5-2b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-5-2b&quant=Q4_K_M&ctx=4096" }, { "id": "qwen3-5-0.8b", "name": "Qwen3.5 0.8B", "family": "Alibaba Qwen3.5", "paramsB": 0.8, "paramLabel": "0.8B", "contextLength": 262144, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-10-11", "arch": { "layers": 24, "kvHeads": 2, "headDim": 256, "attention": { "fullLayers": 6 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 0.49, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 0.85, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/qwen3-5-0.8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=qwen3-5-0.8b&quant=Q4_K_M&ctx=4096" }, { "id": "mimo-v2-6-pro", "name": "MiMo-V2.6-Pro", "family": "Xiaomi MiMo", "paramsB": 1022, "paramLabel": "1T-A42B", "activeParamsB": 42, "contextLength": 1048576, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-12", "arch": { "layers": 70, "kvHeads": 8, "headDim": 160, "attention": { "fullLayers": 10, "windowLayers": 60, "windowTokens": 768 } }, "referenceLevel": "MXFP4", "quants": [ { "format": "GGUF", "level": "MXFP4", "bpw": 4.52, "weightsGB": 577.43, "confidence": "estimated", "pplLossPercent": 0 } ], "url": "https://quantized.uk/quant-hub/mimo-v2-6-pro/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=mimo-v2-6-pro&quant=MXFP4&ctx=4096" }, { "id": "mellum2-1-12b-a2-5b", "name": "Mellum2.1 12B-A2.5B Thinking", "family": "JetBrains Mellum", "paramsB": 12.15, "paramLabel": "12B-A2.5B", "activeParamsB": 2.5, "contextLength": 131072, "categories": [ "code", "instruct" ], "status": "active", "addedToIndex": "2026-10-13", "arch": { "layers": 28, "kvHeads": 4, "headDim": 128, "attention": { "fullLayers": 7, "windowLayers": 21, "windowTokens": 1536 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 5.33, "weightsGB": 8.09, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.49, "weightsGB": 12.89, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/mellum2-1-12b-a2-5b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=mellum2-1-12b-a2-5b&quant=Q4_K_M&ctx=4096" }, { "id": "granite-4.2-8b", "name": "Granite 4.2 8B", "family": "IBM Granite", "paramsB": 8.79, "paramLabel": "8B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-10-14", "arch": { "layers": 40, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.87, "weightsGB": 5.35, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 9.34, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/granite-4.2-8b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=granite-4.2-8b&quant=Q4_K_M&ctx=4096" }, { "id": "granite-4.2-30b", "name": "Granite 4.2 30B", "family": "IBM Granite", "paramsB": 29.28, "paramLabel": "30B", "contextLength": 131072, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-10-16", "arch": { "layers": 64, "kvHeads": 8, "headDim": 128 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.84, "weightsGB": 17.71, "confidence": "estimated" }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.68, "weightsGB": 20.79, "confidence": "estimated" }, { "format": "GGUF", "level": "Q6_K", "bpw": 6.56, "weightsGB": 24.01, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 31.11, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/granite-4.2-30b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=granite-4.2-30b&quant=Q4_K_M&ctx=4096" }, { "id": "granite-4.2-3b", "name": "Granite 4.2 3B", "family": "IBM Granite", "paramsB": 3.66, "paramLabel": "3B", "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-10-16", "arch": { "layers": 40, "kvHeads": 8, "headDim": 64 }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.9, "weightsGB": 2.24, "confidence": "estimated" }, { "format": "GGUF", "level": "Q5_K_M", "bpw": 5.71, "weightsGB": 2.61, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 3.89, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/granite-4.2-3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=granite-4.2-3b&quant=Q4_K_M&ctx=4096" }, { "id": "lfm2-5-24b-a2b", "name": "LFM2.5 24B-A2B", "family": "Liquid AI LFM2.5", "paramsB": 23.97, "paramLabel": "24B-A2B", "activeParamsB": 2.45, "contextLength": 131072, "categories": [ "general", "instruct" ], "status": "active", "addedToIndex": "2026-10-14", "arch": { "layers": 40, "kvHeads": 8, "headDim": 64, "attention": { "fullLayers": 10 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.85, "weightsGB": 14.53, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.51, "weightsGB": 25.5, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/lfm2-5-24b-a2b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=lfm2-5-24b-a2b&quant=Q4_K_M&ctx=4096" }, { "id": "muse-spark-open-816b-a42b", "name": "Muse Spark Open 816B-A42B", "family": "Meta Muse", "paramsB": 820, "paramLabel": "816B-A42B", "activeParamsB": 42, "contextLength": 1048576, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-16", "arch": { "layers": 62, "kvHeads": 16, "headDim": 64, "attention": { "fullLayers": 16, "windowLayers": 46, "windowTokens": 2560 } }, "referenceLevel": "Q4_K_S", "quants": [ { "format": "GGUF", "level": "Q2_K", "bpw": 2.97, "weightsGB": 304.43, "confidence": "estimated" }, { "format": "GGUF", "level": "Q4_K_S", "bpw": 4.59, "weightsGB": 470.48, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.51, "weightsGB": 872.28, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/muse-spark-open-816b-a42b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=muse-spark-open-816b-a42b&quant=Q4_K_S&ctx=4096" }, { "id": "nemotron-3.5-bolt-74b-a9b", "name": "Nemotron 3.5 Bolt 74B-A9B", "family": "NVIDIA Nemotron", "paramsB": 74.04, "paramLabel": "74B-A9B", "activeParamsB": 9, "contextLength": 262144, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-16", "arch": { "layers": 88, "kvHeads": 2, "headDim": 128, "attention": { "fullLayers": 8 } }, "referenceLevel": "IQ2_XXS", "quants": [ { "format": "GGUF", "level": "IQ2_XXS", "bpw": 2.38, "weightsGB": 22.03, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/nemotron-3.5-bolt-74b-a9b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=nemotron-3.5-bolt-74b-a9b&quant=IQ2_XXS&ctx=4096" }, { "id": "nemotron-3.5-super-vl-120b-a12b", "name": "Nemotron 3.5 Super VL 120B-A12B", "family": "NVIDIA Nemotron", "paramsB": 120.76, "paramLabel": "120B-A12B", "activeParamsB": 12, "contextLength": 1048576, "categories": [ "general", "instruct", "code", "multimodal" ], "status": "active", "addedToIndex": "2026-10-16", "arch": { "layers": 88, "kvHeads": 2, "headDim": 128, "attention": { "fullLayers": 8 } }, "referenceLevel": "Q4_0", "quants": [ { "format": "GGUF", "level": "Q4_0", "bpw": 4.76, "weightsGB": 71.85, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.51, "weightsGB": 128.46, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/nemotron-3.5-super-vl-120b-a12b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=nemotron-3.5-super-vl-120b-a12b&quant=Q4_0&ctx=4096" }, { "id": "nemotron-3.5-lightning-30b-a3b", "name": "Nemotron 3.5 Lightning 30B-A3B", "family": "NVIDIA Nemotron", "paramsB": 31.59, "paramLabel": "30B-A3B", "activeParamsB": 3, "contextLength": 262144, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-10-16", "arch": { "layers": 52, "kvHeads": 2, "headDim": 128, "attention": { "fullLayers": 6 } }, "referenceLevel": "Q4_0", "quants": [ { "format": "GGUF", "level": "Q4_0", "bpw": 4.79, "weightsGB": 18.91, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.51, "weightsGB": 33.6, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/nemotron-3.5-lightning-30b-a3b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=nemotron-3.5-lightning-30b-a3b&quant=Q4_0&ctx=4096" }, { "id": "ling-3.1-flash", "name": "Ling-3.1-flash", "family": "inclusionAI Ling", "paramsB": 563.95, "paramLabel": "560B-A25B", "activeParamsB": 25, "contextLength": 262144, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-10-17", "arch": { "layers": 72, "kvHeads": 1, "headDim": 288, "attention": { "fullLayers": 9 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 5.03, "weightsGB": 354.58, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.52, "weightsGB": 600.61, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/ling-3.1-flash/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=ling-3.1-flash&quant=Q4_K_M&ctx=4096" }, { "id": "beam-1.0-501b-a23b", "name": "Beam 1.0 501B-A23B", "family": "Reflection Beam", "paramsB": 501.27, "paramLabel": "501B-A23B", "activeParamsB": 23, "contextLength": 1048576, "categories": [ "general", "instruct", "code" ], "status": "active", "addedToIndex": "2026-10-17", "arch": { "layers": 52, "kvHeads": 8, "headDim": 128, "attention": { "fullLayers": 9, "windowLayers": 43, "windowTokens": 1024 } }, "referenceLevel": "Q8_0", "quants": [ { "format": "GGUF", "level": "Q2_K", "bpw": 2.91, "weightsGB": 182.34, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.5, "weightsGB": 532.6, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/beam-1.0-501b-a23b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=beam-1.0-501b-a23b&quant=Q8_0&ctx=4096" }, { "id": "gemma-4-12b", "name": "Gemma 4 12B IT", "family": "Google Gemma 4", "paramsB": 11.91, "paramLabel": "12B", "contextLength": 262144, "categories": [ "general", "instruct", "multimodal" ], "status": "active", "addedToIndex": "2026-10-19", "arch": { "layers": 48, "kvHeads": 8, "headDim": 256, "attention": { "fullLayers": 8, "windowLayers": 40, "windowTokens": 1536, "fullKvHeads": 1, "fullHeadDim": 512 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.78, "weightsGB": 7.12, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.51, "weightsGB": 12.67, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/gemma-4-12b/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=gemma-4-12b&quant=Q4_K_M&ctx=4096" }, { "id": "north-mini-code-1.0", "name": "North Mini Code 1.0 30B-A3B", "family": "Cohere North", "paramsB": 30.48, "paramLabel": "30B-A3B", "activeParamsB": 3, "contextLength": 262144, "categories": [ "code", "instruct" ], "status": "active", "addedToIndex": "2026-10-20", "arch": { "layers": 49, "kvHeads": 4, "headDim": 128, "attention": { "fullLayers": 13, "windowLayers": 36, "windowTokens": 4608 } }, "referenceLevel": "Q4_K_M", "quants": [ { "format": "GGUF", "level": "Q3_K_M", "bpw": 3.73, "weightsGB": 14.21, "confidence": "estimated" }, { "format": "GGUF", "level": "Q4_K_M", "bpw": 4.92, "weightsGB": 18.75, "confidence": "estimated" }, { "format": "GGUF", "level": "Q8_0", "bpw": 8.51, "weightsGB": 32.42, "confidence": "estimated" } ], "url": "https://quantized.uk/quant-hub/north-mini-code-1.0/", "calculatorUrl": "https://quantized.uk/tools/vram-calc/?mode=forward&model=north-mini-code-1.0&quant=Q4_K_M&ctx=4096" } ] }