← Back to hard AIs

Verify critical details — pricing, licensing, availability — with the model's source before business decisions. Full methodology →

Models · Mistral AI

Ministral 3 8B Instruct

Model family: ministral-3

Balanced mid-size Ministral 3 — 8B parameters with vision, multilingual, 256K context. Fits in 12GB VRAMThe memory built into a GPU. VRAM size determines what models you can load and run — a model's weights must fit in VRAM (or be cleverly swapped in and out). A 7B model in 4-bit quantization needs about 6GB of VRAM; a 70B model in 4-bit needs about 40GB; full-precision frontier models need multiple high-end GPUs. When people talk about a model "fitting" on a GPU, they mean VRAM. in FP8. Apache 2.0.

Identity

Creator
Mistral AI
Model family
ministral-3
Release date
2025-12-01

Technical specs

Parameter count
8B
Context window
262K tokens
Modalities
  • Image Input
  • Text
Primary capabilities
  • Chat
  • Function Calling
  • Instruction Following
  • Long Context
  • Multilingual
  • Tool Use
  • Vision

License

License
Apache 2.0
Commercial use
  • Allowed
Terms
  • Modification
  • Redistribution
  • Attribution

Access

Openness
  • Open Weight
Access methods
  • Api First Party
  • Local Runtime Llama Cpp
  • Local Runtime Lm Studio
  • Local Runtime Ollama
  • Local Runtime Vllm
  • Weights Download Direct
  • Weights Download Hf
Cost tier
  • Mixed

Full model card →