Llama 4 Maverick 17B 128E Instruct — Hardware Requirements & GPU Compatibility
Vision
Llama 4 Maverick is Meta's instruction-tuned, natively multimodal mixture-of-experts model, with roughly 401.6 billion total parameters and about 17 billion active per token routed across 128 experts. It accepts multilingual text and image input and produces multilingual text and code output, using early fusion to integrate vision and language from pretraining rather than bolting on a separate vision encoder; Meta trained it on roughly 22 trillion tokens of multimodal data with an August 2024 knowledge cutoff. It is built for assistant-style chat and visual reasoning tasks. Despite its relatively small active-parameter count, its total size means it needs a multi-GPU server-class setup to run even once quantized. Context length is up to 1,000,000 tokens. It is released under the Llama 4 Community License, a custom license that is free for most commercial and research use but requires companies with more than 700 million monthly active users to request separate permission from Meta. It was published in April 2025.
Llama 4 Maverick 17B 128E Instruct runs on a 512 GB+ Mac or a datacenter GPU
Best local pick → Mac Studio (M3 Ultra, 512GB) · Q4_K_M · 272.7 GB · ~34 tok/s
Macs from 192 GB (IQ3_XXS · ~44 tok/s)
Score 87.1#78 of 106 open#74 of 102 that fit ≤512 GBOther
Not set up for it?
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
llmrun Composite Score →
87.1
- 90% range 84.1–90.1
- #78 of 106 open models
- #74 of 102 that fit ≤512 GB
Run Llama 4 Maverick 17B 128E Instruct
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
ollama run llama4:128x17bQ4_K_M runs on one Instinct MI350X 288 GB · 16K context
Download: Q4_K_M/Llama-4-Maverick-17B-128E-Instruct-Q4_K_M-00001-of-00005.gguf242.77 GB + 4 partsunsloth
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on B300 (288 GB) (15 GB headroom) · Q4_K_M
- Generation speed
- ~208 tok/s
- generation speed
- Cost per 1M output tokens
- $9.25
- per 1M output tokens
How Much VRAM Does Llama 4 Maverick 17B 128E Instruct Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | Runs on | |
|---|---|---|---|---|
| 187.7 GB | 204.2 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Runs on Instinct MI325X 256 GB | |
| 193.3 GB | 209.7 GB | 3-bit small — 3-bit small quantization | Runs on Instinct MI325X 256 GB | |
| 215.3 GB | 231.8 GB | 3-bit medium — 3-bit medium quantization | Runs on Instinct MI325X 256 GB | |
| 220.9 GB | 237.3 GB | 4-bit · legacy — 4-bit legacy quantization | Runs on Instinct MI325X 256 GB | |
| 265.1 GB | 281.5 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Runs on Instinct MI350X 288 GB | |
| 314.7 GB | 331.2 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Runs on 5× H100 SXM 80 GB | |
| 364.4 GB | 380.9 GB | 6-bit, very good — 6-bit quantization, very good quality | Runs on 6× H100 SXM 80 GB | |
| 441.7 GB | 458.2 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Runs on 7× H100 SXM 80 GB |
Selected Q4_K_M
4-bit medium quantization272.7GBat 16K context
- Weights
- 263.7 GB
- KV cache
- 8.8 GB
- Overhead
- 0.3 GB
Server-class: ~272.7 GB at 16K context · one Instinct MI350X 288 GB
Or .
Which Hardware Can Run Llama 4 Maverick 17B 128E Instruct?
Q4_K_M · 265.1 GBLlama 4 Maverick 17B 128E Instruct (Q4_K_M) requires 265.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 345+ GB is recommended. No consumer GPU has enough memory. 1 device with unified memory can run Llama 4 Maverick 17B 128E Instruct, including Mac Studio (M3 Ultra, 512GB).
192+ GB
Q4_K_M · best fit IQ3_XXS (178.9 GB)
6 GPUs · 1 Mac class · 2 device families
B200~248Simulate output on B200B300~248Simulate output on B300Instinct MI350X~90Simulate output on Instinct MI350XM3 Ultra256–512 GB~45Simulate output on Mac Studio (M3 Ultra, 256GB)8x H100 SXM5~327Simulate output on NVIDIA DGX H1008x A100 SXM4~301Simulate output on NVIDIA DGX A100 640GB
By VRAM tier · typical GPUs and Macs at Q4_K_M
Consumer GPUs (up to 32 GB) and Macs up to 128 GB: too small for this model (needs ≥ 129.2 GB)8 GB tier →12 GB tier →16 GB tier →24 GB tier →Macs →
How good is Llama 4 Maverick 17B 128E Instruct?
The llmrun Score merges 15 public benchmark boards into a single figure, where 100 represents the average for open models.
Score vs VRAM in the ≤512 GB tier
- Llama 4 Maverick 17B 128E Instruct
- Other models
- Best Score for its VRAM
#74 of 102 that fit ≤512 GB
Each dot is a model at the best quant that fits ≤512 GB. The line steps up to the best Score available at each amount of VRAM; a dot below it is beaten by a model that needs no more memory.
Full ≤512 GB ranking →Benchmarks
The 6 boards that weigh most in the llmrun Score, of 17 with a result for this model; badges show who ran the test.
- GPQA Diamond
Epoch AI · Apr 2025Independent
Independent67.0%
#38/100 open
- LMArena Text
LMArena · Oct 2026Arena
Arena1327.0
#90/188 open
- AIME 2024/2025
Epoch AI · Apr 2025Independent
Independent20.6%
#56/88 open
- Humanity's Last Exam
Epoch AI · Apr 2025Lab-reported
Lab-reported5.7%
#4/4 open
- ARC-AGI-2
Epoch AI · Apr 2025Independent
Independent0.0%
#20/22 open
- MMLU-Pro
TIGER-LabIndependent
Independent80.5%
#24/146 open
In the ≤512 GB class
102 models fit ≤512 GBSame Score, only models that fit ≤512 GB at their best quant: this model's peers by memory need.
- #1
DeepSeek V4.1 Flash 1017DeepSeek
134.4
458.4 GB
- #2
MiMo V2.6 Pro RLXiaomi
133.3
499.3 GB
- #3
MiMo V2.6 Flash RLXiaomi
131.9
186.5 GB
- #4
GLM 5.3Z.ai
130.1
453.8 GB
- #5
DeepSeek V4.1 FlashDeepSeek
129.4
458.4 GB
- #6
Step 5 PreviewStepFun
128.0
407.5 GB
Official models with an llmrun Score, each answering one question; sizes are each model's default quant (usually Q4_K_M) with a 16K-token context.
Best alternative at this size
GLM 5.3 Flash
Z.ai
- 37 higher Score
- 67.6 GB less VRAM
Score 124.1 · 205.2 GB
Other vendor
DeepSeek V4.1 Flash 1017
DeepSeek
- 47.3 higher Score
- 185.7 GB more VRAM
Score 134.4 · 458.4 GB
Similar models
10| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| MiMo V2.6 Flash RLXiaomi | 131.9 | 186.5 GB | —needs > 24 GB | similar score |
| GLM 5.3Z.ai | 130.1 | 453.8 GB | —needs > 24 GB | similar score |
| DeepSeek V4.1 FlashDeepSeek | 129.4 | 458.4 GB | —needs > 24 GB | similar score |
| Step 5 PreviewStepFun | 128.0 | 407.5 GB | —needs > 24 GB | similar score |
| DeepSeek V4 Flash 0731DeepSeek | 127.6 | 183 GB | —needs > 24 GB | similar score |
| GLM 5.2Z.ai | 122.9 | 453.8 GB | —needs > 24 GB | similar score |
| DeepSeek V3.1DeepSeek | 100.2 | 412.2 GB | —needs > 24 GB | similar score |
| Qwen3 235B A22BAlibaba | 100.0 | 142.9 GB | —needs > 24 GB | similar score |
| DeepSeek R1DeepSeek | 98.9 | 412.2 GB | —needs > 24 GB | similar score |
| Qwen3 235B A22B Instruct 2507Alibaba | 98.5 | 142.9 GB | —needs > 24 GB | similar score |
Specifications
- Publisher
- Meta
- Family
- Llama 4
- Parameters
- 401.6B
- Active parameters
- 17B per token
- Release Date
- 2025-04-01
- License
- Other
Frequently Asked Questions
How much VRAM does Llama 4 Maverick 17B 128E Instruct need?
Llama 4 Maverick 17B 128E Instruct requires 265.1 GB of VRAM at Q4_K_M, or 883.5 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 401.6B × 4.8 bits ÷ 8 = 241 GB
KV Cache + Overhead ≈ 24.1 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
VRAM usage by quantization
Can NVIDIA GeForce RTX 5090 run Llama 4 Maverick 17B 128E Instruct?
No — Llama 4 Maverick 17B 128E Instruct requires at least 121.5 GB at IQ2_XXS, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
What's the best quantization for Llama 4 Maverick 17B 128E Instruct?
For Llama 4 Maverick 17B 128E Instruct, Q4_K_M (265.1 GB) offers the best balance of quality and VRAM usage. Q5_K_S (303.7 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 121.5 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Llama 4 Maverick 17B 128E Instruct on a Mac?
Yes — Mac Pro M2 Ultra (192 GB) and 3 other Macs can run Llama 4 Maverick 17B 128E Instruct, the smallest at IQ3_XXS. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 265.1 GB of usable unified memory (RAM minus macOS overhead), from Mac Studio (M3 Ultra, 512GB).
Can I run Llama 4 Maverick 17B 128E Instruct locally?
Yes — Llama 4 Maverick 17B 128E Instruct can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Llama 4 Maverick 17B 128E Instruct?
At Q4_K_M, Llama 4 Maverick 17B 128E Instruct can reach ~84 tok/s on AMD Instinct MI350X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)
Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.
Example: NVIDIA B300 → 11.2 GB active ÷ (8000 × 0.65) = 2.16 ms, plus 48 layers × 0.055 ms = 2.64 ms, so 1000 ÷ 4.80 ms = ~208 tok/s
Estimated speed at Q4_K_M (265.1 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Llama 4 Maverick 17B 128E Instruct?
At Q4_K_M, the download is about 240.95 GB. The full-precision BF16 version is 803.17 GB. The smallest option (IQ2_XXS) is 110.44 GB.
Which GPUs can run Llama 4 Maverick 17B 128E Instruct?
No single consumer GPU has enough VRAM to run Llama 4 Maverick 17B 128E Instruct at Q4_K_M (265.1 GB). Multi-GPU or professional hardware is required.
Which devices can run Llama 4 Maverick 17B 128E Instruct?
1 device with unified memory can run Llama 4 Maverick 17B 128E Instruct at Q4_K_M (265.1 GB), including Mac Studio (M3 Ultra, 512GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.