Model

Mistral Small 3 (24B)

OPEN

Mistral AI · Mistral Small family · released Jan 30, 2025

24B Apache-2.0 latency-optimized model competitive with Llama 3.3 70B at 3x+ the speed; 81% MMLU.

ReasoningCodingVisionFunction callingTool useAgentic

978.8

Elo · rank #217

Parameters

24B

Active params

24B (dense)

Context

32K tokens

Architecture

Dense transformer — 24B, Tekken tokenizer (131k vocab)

License

Apache 2.0

Languages

—

API price (in/out)

$0.1 / $0.3

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

Arena-HardHuman preference87.3%#18

best: Qwen3-235B-A22B (Non-Thinking) · 96.1%

HumanEvalCoding84.8%#44

best: Claude Opus 4.5 · 99.4%

IFEvalReasoning82.9%#84

best: Gemma 4 26B A4B · 98.5%

MATH-500Math70.6%#88

best: GPT-5 · 99.4%

MMLU-ProKnowledge66.3%#103

best: Claude Fable 5 · 91.5%

MMLUKnowledge81.0%#83

best: OpenAI o3 · 92.9%

MT-BenchHuman preference8.35#36

best: Hunyuan-Large (A52B) · 9.4

WildBenchHuman preference52.3%#4

best: Mistral Small 3.2 (24B) · 65.3%

Run it locally

VRAM @ Q4

14.5 GB

VRAM @ FP16

55 GB

Fits on (Q4)

RTX 4070 Ti 16GBRTX 3090 24GBRTX 4090 24GBRTX 5090 32GBM4 Pro 48GBM3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB

Throughput data unavailable.

Quantizations

GGUF Q4 · AWQ · GPTQ · MLX

Mistral Small family

Elo progression across releases

Mistral Small 3 (24B)Jan 2025978.8 Mistral Small 3.1 (24B)Mar 20251051.4 Mistral Small 3.2 (24B)Jun 20251075.6

Resources

HFmistralai/Mistral-Small-24B-Instruct-2501 GHmistralai/mistral-inference DOCSdocs.mistral.ai/models/mistral-small-3-0-25-01

API price $0.1/$0.3 · each benchmark row carries its own source badge (see methodology)