How much VRAM does Llama 4 Maverick 17B (128E) need?

About 254.2 GB at Q4_K_M with 4K of context and batch 1, rising to roughly 256.4 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 Llama 4 Maverick 17B (128E) run on a 512GB GPU?

Yes — at Q4_K_M and 4K context it needs about 254.2 GB, which leaves 129.8 GB spare on a Mac M5 Ultra 512G. That is the smallest card in this index that clears it comfortably; 1 of 78 do.

Which quantization of Llama 4 Maverick 17B (128E) should I use?

Q4_K_M has the lowest published quality loss (2.2%), and Q4_K_M is the level most people run. All 2 levels in the index are Q4_K_M, Q3_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.

Llama 4 Maverick 17B (128E)

400B MoE

Meta Llama 4

Llama 4 Maverick flagship MoE (17B active / 400B total). Multi-GPU or H100 cluster territory.

⬇ 11.9K HF downloads♥ 51 likesunsloth/Llama-4-Maverick-17B-128E-Instruct-GGUF· stats from 10/20/2026

Pro GPU

1049K

Max Context

Quant Variants

GGUF Q4_K_M

Best Quality

97.8%

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	Q4_K_M	4.85	245.0 GB	2.2%	—	Estimated	Calc HF
GGUF	Q3_K_M	3.87	198.0 GB	4.5%	—	Estimated	Calc HF

Similar models

Compare with Llama 4

109B MoE

Llama 4 Scout 17B (16E)

Meta Llama 4

Pro GPU

55.0 GBmin VRAM97.6%accuracy

Meta Llama 4 Scout MoE (17B active / 109B total). Multimodal; needs ~68GB VRAM at Q4_K_M.

675B MoE

Mistral Large 3 675B Instruct

Mistral AI

Pro GPU

312.0 GBmin VRAM97.9%accuracy

Mistral 3 flagship MoE (41B active / 675B total) with vision encoder. FP8 on 8×H200; GGUF quant for research clusters only. Its attention is DeepSeek-style MLA, so the real KV cache is smaller than the figure below, which uses an unverified conventional layout. At 32K that is a rounding error beside ~390 GB of weights; at the full 256K window, treat the cache figure as an upper bound.

1T-A42B

MiMo-V2.6-Pro

Xiaomi MiMo

Pro GPU

603.8 GBmin VRAM100.0%accuracy

Xiaomi's open-weight flagship, near the top of the open-model rankings this autumn. It has about 1.02 trillion parameters, roughly 42B of them active per token, and ships under the MIT licence. The released weights are already 4-bit (MXFP4 experts), and the GGUF builds repack them rather than quantize again. At that level it needs about 604 GB at 4K context, more than any single machine in this index. The largest Mac Studio gives its GPU about 384 GB, and four 80 GB H100s hold 320 GB. Eight 80 GB cards (640 GB) are about the smallest server that holds it, and even that is tight on this site's rule. It is listed so the calculator can say so. Only 10 of its 70 layers keep a growing cache, so even 1M tokens adds only about 50 GB. llama.cpp runs it as `mimo2`, and vision loads through a separate encoder. Sizes come from a published GGUF file; 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 Llama 4 Maverick 17B (128E) locally

At Q4_K_M and 4K of context, Llama 4 Maverick 17B (128E) needs about 254.2 GB — 230.4 GB of weights, 0.75 GB of KV cache and a 23.1 GB activation buffer. The smallest card in this index that clears that comfortably is the Mac M5 Ultra 512G at 512 GB, and 1 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 1.95 GB, taking the total to 256.4 GB. That is gentler than the parameter count suggests: only 12 of 48 layers keep a KV cache that grows with context, and 36 more cap theirs at a sliding window, so the cache scales at a fraction of the usual rate. The model's native window is 1M; holding all of it at Q4_K_M would need about 308 GB.

Which build to download

This index only tracks Llama 4 Maverick 17B (128E) in GGUF, across 2 levels (Q4_K_M, Q3_K_M). Q4_K_M carries the lowest published perplexity loss at 2.2%. 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 Llama 4 Maverick 17B (128E) need?: About 254.2 GB at Q4_K_M with 4K of context and batch 1, rising to roughly 256.4 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 Llama 4 Maverick 17B (128E) run on a 512GB GPU?: Yes — at Q4_K_M and 4K context it needs about 254.2 GB, which leaves 129.8 GB spare on a Mac M5 Ultra 512G. That is the smallest card in this index that clears it comfortably; 1 of 78 do.
Which quantization of Llama 4 Maverick 17B (128E) should I use?: Q4_K_M has the lowest published quality loss (2.2%), and Q4_K_M is the level most people run. All 2 levels in the index are Q4_K_M, Q3_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 Q4_K_M with a 4K context window, smallest card first. Comfortable means the estimate uses at most 88% of the memory.

Runs comfortably on: Mac M5 Ultra 512G512GB · +129.8
Just misses: Mac M5 Ultra 256G+62.2GB over Mac M2 Ultra 192G+110.2GB 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