Modelofficial release

mistralai/Devstral-2-123B-Instruct-2512

Mistral models →Hugging Face ↗
Parameters125B
Context256K
Input → output–
Licenseother
PrecisionF8_E4M3 · fp8
ArchitectureMinistral3ForCausalLM
ReleasedNov 28, 2025
Downloads5.2K

This release in the news

No coverage of this exact release yet.

Recent news about Mistral models

  1. Oct 17, 2026 · Tutorial / explainer
    Local and Distributed Large LLM Inference with MLX and RDMA on Mac Studio
  2. Oct 17, 2026 · Opinion / analysis
    llama.cpp has come a long way
  3. Oct 16, 2026 · Research paper
    Recognizing Study Lineage Is Not Enough: Over-Crediting Same-Trial Evidence in LLM-Based Medical Verification
  4. Oct 15, 2026 · Opinion / analysis
    I benchmarked whether ASR captures a dictated phone number exactly, in German and French. Whisper: 34% exact. Same text written down: 91% [P]
  5. Oct 14, 2026 · Research paper
    SpectrumQuant: Spectrum Alignment for LLM Quantization
  6. Oct 14, 2026 · Research paper
    Beyond the Trial-and-Error Loop: Hybrid Projection and Automated Tuning for Distributed Training
  7. Oct 13, 2026 · Research paper
    Sharing AI progress in mathematics
  8. Oct 13, 2026 · Research paper
    ATTUNE: Agentic Throughput Tuning for LLM Training