Plain-English Summary
A production model balancing strong agentic and coding performance with lower cost than Mistral Large.
Mistral Medium 3.5
A production model balancing strong agentic and coding performance with lower cost than Mistral Large.
Mistral Medium
Multimodal LLM
GA / open weights
Open weights: Yes
API: Yes and self-hosted
Technical Notes
128B dense multimodal model with 256K context, text/image input, tool use, and enterprise deployment options.
Best For
Enterprise agents, coding, document analysis, multimodal RAG
Watch Out For
Dense 128B deployment remains demanding; modified license differs from Apache 2.0.
Model Specs
- Maker
- Mistral AI
- Country
- France
- Family
- Mistral Medium
- Type
- Multimodal LLM
- Release date
- 2026-04-28
- Date confidence
- Exact
- Status
- GA / open weights
- Context window
- 262,144 tokens
- Max output
- Not listed
- Parameters
- 128B
- Active parameters
- 128B
- Architecture
- Dense multimodal transformer
- Reasoning
- Yes
- Tool calling
- Yes
- Structured output
- Yes
- API available
- Yes and self-hosted
- Open weights
- Yes
- Self-hostable
- Yes
- License
- Modified MIT
- Inputs
- Text, image
- Outputs
- Text
- Approx. price
- $1.50 input / $7.50 output per 1M tokens
- How to access
- Provider website
- Model / API ID
- mistral-medium-3-5-26-04
- Knowledge cutoff
- Not publicly disclosed
- Fine-tuning
- Fine-tuning, self-hosting, tools
- Completeness
- 95 %
- Verified
- Jul 18, 2026
- Snapshot date
- Jul 18, 2026
- Source
- Official source