Ministral 3 14B
Ministral 3 14B is an open-weight language model from Mistral AI, released on December 2, 2025. It is a dense model with 14B parameters. That figure covers a 13.5B language model and a 0.4B vision encoder. Mistral AI lists a context window of 256K tokens. The model accepts text and image input.
- Developer
- Mistral AI
- Released
- December 2, 2025
- Total parameters
- 14B
- Architecture
- Dense
- Context window
- 256K
- Input
- Text, image
- License
- Apache 2.0
- Commercial use
- Yes
- Official weights, GB
- 15.7 GB
- Hugging Face repository
- huggingface.co
- Checked on
The weights are released under Apache 2.0, which allows commercial use.
The official weights on Hugging Face take about 15.7 GB in FP8. Running the model needs at least that much memory across GPUs and system RAM, plus room for the context cache; quantized versions need less.
- Total parameters, billions
- 14B
- Context window, tokens
- 256K
- License type
- Apache 2.0
- License conditions
- Apache 2.0, with no user or revenue limits.
- Weights precision
- FP8
- Official quantized or alternative versions
- mistralai/Ministral-3-14B-Instruct-2512-GGUF; mistralai/Ministral-3-14B-Instruct-2512-BF16 (full precision)
- License text
- apache.org
- Sources
Page URL Hugging Face model card https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512/raw/main/README.md Hugging Face weights index https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512/raw/main/model.safetensors.index.json mistral.ai announcement https://mistral.ai/news/mistral-3