VOOZH about

URL: https://willitrunai.com/models/magistral-small-2507

โ‡ฑ Magistral Small 2507 VRAM Requirements โ€” GPU Compatibility


๐Ÿ‘ Mistral
Mistral

Magistral Small 2507

Legacy
๐Ÿ‘ huggingface
HuggingFace๐Ÿ‘ ollama
Ollama
1.0KDownloads104LikesJul 2025Released131K tokensContextApache 2.0License96 ExceptionalQuality

Magistral Small 2507 (24B parameters) requires approximately 18.9 GB of VRAM with Q4_K_M quantization. For the best balance of quality and speed, we recommend hardware with at least 22 GB of VRAM.

Get started

โ€” copy & paste to run locally

Copy-paste commands to run Magistral Small 2507 on your machine.

Run

ollama run magistral

Quick specs

Parameters24B
Architecturedense
Context131K tokens
Modalitytext
Min RAM9.4 GB
Rec. RAM14.6 GB (Q4_K_M)
LicenseApache 2.0
FamilyMagistral
โœ“ Chatโœ“ Reasoning

About this model

Building upon Mistral Small 3.1 (2503), with added reasoning capabilities, undergoing SFT from Magistral Medium traces and RL on top, it's a small, efficient reasoning model with 24B parameters.

  • โ€ขReasoning:: Capable of long chains of reasoning traces before providing an answer
  • โ€ขMultilingual:: Supports dozens of languages, including English, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay,...
  • โ€ขApache 2.0 License:: Open license allowing usage and modification for both commercial and non-commercial purposes
  • โ€ขContext Window:: A 128k context window, but performance might degrade past 40k. Hence we recommend setting the maximum model length to 40k

Related models

Your hardware

Detecting...

Quick picks

๐Ÿ‘ Intel
Best budgetS
Intel Arc Pro B60 24GB~$599 โ€” 18 tok/s
๐Ÿ‘ NVIDIA
Best overallS
RTX 5090 32GB~$1,999 โ€” 88 tok/s

Best hardware

Top picks for Magistral Small 2507

RTX 5090 32GBS
32 GB
RTX PRO 4500 Blackwell 32GBS
32 GB
AMD Instinct MI100 32GBS
32 GB
NVIDIA V100 32GBS
32 GB
NVIDIA A100 40GBS
40 GB

Run this model

Magistral Small 2507 on RTX 5090 32GBMagistral Small 2507 on RTX PRO 4500 Blackwell 32GBMagistral Small 2507 on AMD Instinct MI100 32GB

Quantization options

VRAM estimates by quant level

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

QuantBitsVRAMQualityFit
Q2_K
2
9.4 GB
Lowโ€”
Q3_K_S
3
11.8 GB
Lowโ€”
NVFP4
4
13.4 GB
Mediumโ€”
Q4_K_M
4
14.6 GB
Mediumโ€”
Q5_K_M
5
17.3 GB
Highโ€”
Q6_K
6
19.7 GB
Highโ€”
Q8_0
8
25.7 GB
Very Highโ€”
F16
16
49.2 GB
Maximumโ€”

Quality benchmarks

Magistral Small 2507 benchmark scores

Benchmark verified

Reasoning

MMLU-Proโ€”
GPQA Diamond68.2%
MATH-50095.9%
ARC Challengeโ€”

General

Chatbot Arenaโ€”
IFEval87.4%

Source: official ยท 2025-07-01

Hardware compatibility

Fit estimates across all hardware

Open calculator

Computing compatibility...

Memory breakdown

Reference: RTX 2060 6GB

Weights14.6 GB
KV Cache2.4 GB
Runtime1.2 GB
Headroom0.6 GB

Frequently asked questions

FAQ โ€” Magistral Small 2507

See also

Quantization GuideScoring MethodologyVRAM Calculator