VOOZH about

URL: https://willitrunai.com/models/mistral-large-3-675b-a41b

โ‡ฑ Mistral Large 3 VRAM Requirements โ€” GPU Compatibility


๐Ÿ‘ Mistral
Mistral

Mistral Large 3

Frontier
๐Ÿ‘ huggingface
HuggingFace
8.3KDownloads262LikesNov 2024Released256K tokensContextMistral Research LicenseLicense75 StrongQuality

Mistral Large 3 (675B parameters) requires approximately 420.1 GB of VRAM with Q4_K_M quantization. As a Mixture of Experts model with 41B active parameters, it uses less memory than its total parameter count suggests. For the best balance of quality and speed, we recommend hardware with at least 484 GB of VRAM.

Get started

โ€” copy & paste to run locally

Quick specs

Parameters675B (41B active)
Architecturemoe (MoE)
Context256K tokens
Modalitytext+vision
Min RAM263.3 GB
Rec. RAM411.8 GB (Q4_K_M)
LicenseMistral Research License
FamilyMistral Large
โœ“ Visionโœ“ Codeโœ“ Chatโœ“ Reasoning

About this model

Mistral-Large-Instruct-2411 is an advanced dense Large Language Model (LLM) of 123B parameters with state-of-the-art reasoning, knowledge and coding capabilities extending Mistral-Large-Instruct-2407 with better Long Context, Function Calling and System Prompt.

  • โ€ขMulti-lingual by design:: Dozens of languages supported, including English, French, German, Spanish, Italian, Chinese, Japanese, Korean,...
  • โ€ขProficient in coding:: Trained on 80+ coding languages such as Python, Java, C, C++, Javacsript, and Bash. Also trained on more specific languages...
  • โ€ขAgent-centric:: Best-in-class agentic capabilities with native function calling and JSON outputting
  • โ€ขAdvanced Reasoning:: State-of-the-art mathematical and reasoning capabilities
  • โ€ขMistral Research License:: Allows usage and modification for non-commercial usages

Your hardware

Detecting...

Quantization options

VRAM estimates by quant level

No hardware detected โ€” fit column shows raw VRAM estimates

QuantBitsVRAMQualityFit
Q2_K
2
263.3 GB
Lowโ€”
Q3_K_S
3
330.8 GB
Lowโ€”
NVFP4
4
378.0 GB
Mediumโ€”
Q4_K_M
4
411.8 GB
Mediumโ€”
Q5_K_M
5
486.0 GB
Highโ€”
Q6_K
6
553.5 GB
Highโ€”
Q8_0
8
722.3 GB
Very Highโ€”
F16
16
1383.7 GB
Maximumโ€”

Quality benchmarks

Mistral Large 3 benchmark scores

Benchmark verified

Coding

SWE-bench Verifiedโ€”
HumanEval+92.0%
Aider Polyglotโ€”
LiveCodeBench34.4%

Reasoning

MMLU-Pro85.5%
GPQA Diamond43.9%
MATH-50049.5%
ARC Challengeโ€”

General

Chatbot Arenaโ€”
IFEval84.0%

Source: official ยท 2025-12-18

Hardware compatibility

Fit estimates across all hardware

Open calculator

Computing compatibility...

Memory breakdown

Reference: RTX 2060 6GB

Weights411.8 GB
KV Cache5.4 GB
Runtime2.4 GB
Headroom0.6 GB

Frequently asked questions

FAQ โ€” Mistral Large 3

See also

Quantization GuideScoring MethodologyVRAM Calculator