release aidev NEEDS REVIEW SIG 1/5

Mistral publishes Devstral-Small-2-24B-Instruct-2512

Mistral released a 24B parameter instruct model on Hugging Face under Apache 2.0, fine-tuned from Mistral-Small-3.1-24B-Base-2503. The model uses the Mistral 3 architecture, ships with FP8 weights, and is served via vLLM. The repository does not declare a pipeline tag or provide capability details.

PUBLISHED2025-11-28
OBSERVED2026-08-11
AGE9mo
SOURCES1
  • 24B parameter instruct model, Mistral 3 architecture, FP8 safetensors
  • Fine-tuned from mistralai/Mistral-Small-3.1-24B-Base-2503
  • Apache 2.0 license, vLLM library
  • No pipeline tag declared; no capability or benchmark information in the repository

COMMUNITY

No curated reactions recorded for this event. Facts and takes are kept in separate layers — community context is added by hand, never blended into the record above.