Mistral Small 24B Instruct 2501

mistralai · released 2025-01-30 · apache-2.0 license

Mistral Small 24B Instruct 2501 (Jan 2025, Apache-2.0) is a knowledge-dense 24B rivaling much larger models: vendor MMLU-Pro 66.3, IFEval 82.9.

Key specs

TypeLocal open-weight
Parameters23.57B total
Architecturemistral
Context window33K tokens
Knowledge cutoff2023-10-31
Modalitiestext
Recommended backends
Minimum viable rig

Benchmark scores

GPQA Diamond
SWE-bench Verified
AIME
MMLU-Pro66.3%
BFCL v3 (tool use)
Composite score7.31
Community ratingNo reviews yet

VRAM & disk per quantization

QuantVRAMDiskRAMContext
Q4_K_M15.2 GB13.7 GB33K

Strengths & weaknesses

Strengths: SOTA for its size class (<70B), rivaling Llama-3.3-70B / Qwen2.5-32B; Strong function-calling and JSON output; Runs on a single 4090; Apache-2.0

Weaknesses: 32K context smaller than many competitors; No built-in safety guardrails