DownloadMinistral 3 8B Reasoning 2512 is a balanced midsize model in the Ministral 3 family, delivering strong multimodal reasoning capabilities within an efficient footprint. It combines an 8.4B-parameter language model with a 0.4B vision encoder, enabling it to process both text and images for advanced reasoning tasks. This version is specifically post-trained for reasoning, making it well-suited for math, coding, and STEM applications requiring multi-step logic and problem-solving. Despite its reasoning-focused training, the model remains edge-optimized and can run locally on a single 24GB GPU in BF16, or under 12GB when quantized. It supports dozens of languages, adheres reliably to system prompts, and provides native function calling and structured JSON output—key capabilities for agentic and automation workflows. The model also includes a 256k context window, allowing it to handle long documents and extended reasoning chains.
Features
- 8.4B language model with a 0.4B vision encoder for multimodal reasoning
- Post-trained specifically for math, coding, and STEM-related reasoning tasks
- Runs locally on 24GB VRAM in BF16 or <12GB with quantization
- Supports dozens of major world languages including English, Spanish, Chinese, Arabic, and more
- Strong system-prompt adherence for predictable reasoning behavior
- Native agentic capabilities with function calling and structured JSON output
- Large 256k context window for long-form and multi-document problem-solving
- Edge-optimized design suitable for local, embedded, and resource-constrained deployments
Project Samples
