magistral:24b

Upstream reference: Magistral Small 2506

Publisher: Mistral AI

Mistral documents this reasoning-oriented release as a derivative of Mistral Small 3.1, producing reasoning text before a final answer.

Publisher-described context, separate from Sigma5C observations and recorded catalog statistics.

Published upstream specifications

Parameters

24 billion Source

Published context

128k tokens, publisher notation Source

Upstream license label

Apache-2.0 Source

Publisher-described uses

• Multilingual reasoning-oriented text generation. Source

Reference limitations

• The publisher notes possible degradation beyond 40k context; the 128k figure should be shown with that material limitation. Source

A recorded tag and an upstream card association do not verify the deployed checkpoint, quantization, configured context or current hosted availability.

Publisher sources

• Magistral Small 2506 official documentation — Mistral AI. Checked 2026-11-02.

Links

Recorded references do not establish downloadable weights, licensing or current availability.

• Mistral AI — official site
• Ollama
• Hugging Face
• Official model announcement / documentation

Where magistral:24b comes from

Named in 2 observation pages

• 14B-32B local models — all 47 · V1
• Thinking-token re-observation — 14B-32B · V2