How much VRAM does GPT-OSS 120B need?

About 66.4 GB at MXFP4 with 4K of context and batch 1, rising to roughly 67.5 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card.

Will GPT-OSS 120B run on a 80GB GPU?

Yes — at MXFP4 and 4K context it needs about 66.4 GB, which leaves 13.6 GB spare on a A100 80G, or on any of the other 80 GB cards here (2 in all). That is the smallest size in this index that clears it comfortably; 10 of 78 do.

Which quantization of GPT-OSS 120B should I use?

MXFP4 has the lowest published quality loss (0.0%), and MXFP4 is the level most people run. All 2 levels in the index are MXFP4, Q4_K_M. Note this is a mixture-of-experts model — all parameters must be resident even though only a fraction are active per token, so the memory cost follows the total, not the active count.

GPT-OSS 120B

117B MoE

OpenAI GPT-OSS

The big GPT-OSS (117B total / 5.1B active). Native MXFP4 checkpoint is ~61GB — fits one 80GB card or a 128GB unified-memory Mac. Partial offload on 24GB consumer cards is slow but works.

⬇ 3.2M HF downloads♥ 5426 likesopenai/gpt-oss-120b· stats from 10/20/2026

Pro GPUMac / Apple Silicon

131K

Max Context

Quant Variants

GGUF MXFP4

Best Quality

100.0%

Accuracy Retained

Calculate VRAM Hugging Face Compare

Quantization Variants

Per-quant VRAM, quality loss, and inference speed on RTX 4090

Measured = run on this site · Estimated = not run here: calculated, or a figure this site has not verified · Community = public reports

Format	Level	BPW	VRAM	PPL Loss	Speed	Source	Actions
GGUF	MXFP4	4.34	61.0 GB	0.0%	—	Community	Calc HF
GGUF	Q4_K_M	4.1	58.5 GB	1.6%	—	Estimated	Calc HF

Similar models

Compare with GPT-OSS 20B

21B MoE

GPT-OSS 20B

OpenAI GPT-OSS

Consumer GPUMac / Apple Silicon

11.9 GBmin VRAM100.0%accuracy

OpenAI open-weight MoE (21B total / 3.6B active), shipped natively in MXFP4 — ~12.8GB runs on a 16GB card with no quality tax. Only 3.6B active params means CPU-offload stays usable.

106B MoE

GLM-4.5-Air

Zhipu GLM-4.5

Pro GPUMac / Apple Silicon

34.8 GBmin VRAM97.6%accuracy

Zhipu's agentic/reasoning MoE (106B total / 12B active). Q4 ~64GB — the sweet spot is a 96GB+ unified Mac or 2× 48GB cards. Strong tool-calling for its class.

125B-A6B

Qwen3.8-Flash-Next

Qwen3.8

Mac / Apple SiliconPro GPU

79.3 GBmin VRAM—accuracy

Qwen's preview of its next architecture: a 125B mixture-of-experts model with about 6B parameters active per token. Released in August, it also carries a 51B n-gram lookup table. llama.cpp reads that table lazily from RAM or disk, so it never needs VRAM. It does need disk space and is bundled inside the GGUF files, which is why they are far larger than the sizes below. Only 12 of its 48 layers keep a growing KV cache, so even the full 256K window adds about 7.5 GB. At Q4_K_M it needs about 79 GB at 4K context and 87 GB at 256K, comfortable on a 128 GB Mac or DGX Spark. A single 24 or 32 GB card runs it only with experts offloaded to system RAM. llama.cpp supports it as `qwen4exp`. Sizes use the calculator's generic rates, since no file size isolates the model from the table, and no per-level quality loss has been published.

816B-A42B

Muse Spark Open 816B-A42B

Meta Muse

Mac / Apple SiliconPro GPU

318.9 GBmin VRAM—accuracy

Meta's largest open-weight model, released in October under the Muse Model License. It is an 816B mixture-of-experts model with about 42B parameters active per token, built for coding and agent work, with image and audio input. It is a single-machine model only at its smallest build. At Q2_K (ggml-org's 304 GB file) it needs about 319 GB at 4K context and still only 327 GB at 128K, because just 16 of its 62 layers keep a growing cache — comfortable on a 512 GB Mac Studio and tight across four 80 GB H100s. The full 1M window brings it to about 389 GB. Q4_K_S (470 GB) and Q8_0 (872 GB) fit no single machine in this index. The vision and audio encoders ship as separate files of 2–4 GB and are not included. Sizes come from the published GGUF files; Meta's benchmark claims are its own and no per-level quality loss has been published. The licence is not a standard open-source one: Meta's repository is gated, so downloads need an accepted access request; products built on it must show "Built with Muse Spark Open"; and a company with over $20M in annual revenue must sign a separate agreement with Meta before offering it as an API service.

Running GPT-OSS 120B locally

At MXFP4 and 4K of context, GPT-OSS 120B needs about 66.4 GB — 60.2 GB of weights, 0.17 GB of KV cache and a 6.0 GB activation buffer. MXFP4 is the format GPT-OSS 120B was released in, so it is the reference here rather than a community requantization. The smallest cards in this index that clear that comfortably are 80 GB ones — 2 of them, from the A100 80G up — and 10 of the 78 cards here do. These are calculated figures, not measurements: the estimate stops counting a card as comfortable at 88% of its VRAM, which is roughly the room a desktop session needs.

What longer context costs

Going from 4K to 32K adds about 0.98 GB, taking the total to 67.5 GB. That is gentler than the parameter count suggests: only 18 of 36 layers keep a KV cache that grows with context, and 18 more cap theirs at a sliding window, so the cache scales at a fraction of the usual rate. The model's native window is 128K; holding all of it at MXFP4 would need about 71 GB.

Which build to download

This index only tracks GPT-OSS 120B in GGUF, across 2 levels (MXFP4, Q4_K_M). MXFP4 carries the lowest published perplexity loss at 0.0%. No throughput has been measured for this model on this site's hardware. GGUF runs on llama.cpp and Ollama across NVIDIA, AMD and Apple silicon; AWQ and GPTQ target vLLM on CUDA and ROCm; EXL2 is ExLlamaV2 — CUDA only, and now archived, so local use rather than serving.

Common questions

How much VRAM does GPT-OSS 120B need?: About 66.4 GB at MXFP4 with 4K of context and batch 1, rising to roughly 67.5 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card.
Will GPT-OSS 120B run on a 80GB GPU?: Yes — at MXFP4 and 4K context it needs about 66.4 GB, which leaves 13.6 GB spare on a A100 80G, or on any of the other 80 GB cards here (2 in all). That is the smallest size in this index that clears it comfortably; 10 of 78 do.
Which quantization of GPT-OSS 120B should I use?: MXFP4 has the lowest published quality loss (0.0%), and MXFP4 is the level most people run. All 2 levels in the index are MXFP4, Q4_K_M. Note this is a mixture-of-experts model — all parameters must be resident even though only a fraction are active per token, so the memory cost follows the total, not the active count.

Where this model fits

Sized at MXFP4 with a 4K context window, smallest card first. Comfortable means the estimate uses at most 88% of the memory.

Runs comfortably on: A100 80G80GB · +13.6 H100 80G80GB · +13.6 DGX Spark 128G128GB · +61.6 Mac M3 Max 128G128GB · +29.6 Mac M4 Max 128G128GB · +29.6
Tight but possible: Mac M2 Max 96G96GB
Just misses: DGX Spark 64G+2.4GB over A40 48G+18.4GB over
Ships in: GGUF
Try it yourself: Size it in the calculator Generate the run command

How to actually run this

Deployment guides for this model and this class of hardware.

This page's figures change when the model or the runtime does.Last updated 2026-10-20 RSS → /feed.xml