Mistral Nemo 12B Instruct 2407

Permissive

Mistral AI / NVIDIA · Mistral · Released 2024-07

12BText

A joint Mistral/NVIDIA release that packs a 128K context window and a new Tekken tokenizer into a single-consumer-GPU-friendly 12B.

Strengths

  • +Apache 2.0, unrestricted use
  • +128K context window
  • +Efficient Tekken tokenizer improves multilingual token efficiency

Limitations

  • -Reasoning trails newer 2025 models at similar size (Qwen2.5-14B)
  • -Less community fine-tuning activity than Llama/Qwen

License

Apache 2.0

use commercially with attribution niceties

Hardware

runs on a good consumer GPU (or Apple Silicon) with quantization

Links

Stats

mistralai/Mistral-Nemo-Instruct-2407— downloads·— likes

via Hugging Face

Related models