# Mistral Large 4

Mistral Large 4 is a state-of-the-art, open-weight, general-purpose multimodal model with a granular Mixture-of-Experts architecture. It features 52B active parameters and 1.05T total parameters, and a 1.6B vision encoder.

- API names: `mistral-large-4`, `mistral-large-4-0`
- Status: Public Preview
- Type: Open, Generalist
- Release date: 2026-10-06
- Version: 26.10
- Context length: 1M
- Parameters: 1.05T
- Input: text, image
- Output: reasoning, text
- Features: structured-outputs, function-calling, document-qna, prefix, chat-completions, batching, agents-conversations, connectors
- Speed rating: 5/5
- Performance rating: 4/5
- Pricing: Input $0.68 /M Tokens (sale price, regular $1.36); Cached input $0.07 /M Tokens (sale price, regular $0.14); Output $2.09 /M Tokens (sale price, regular $4.18)
- Open weights: [Coming soon](https://huggingface.co/mistralai/Mistral-Large-4-1T-A52B), 1.05T parameters, 52B active