Mistral Medium 3.5
Mistral Medium 3.5
Mistral's closed middle tier: multimodal, agent-focused, and priced to be used.
Updated
The read
Medium 3.5 is where Mistral put its agentic and multimodal work, and it is not published as weights. It handles tool calling and coding well, costs meaningfully less than the American frontier models, and comes with the European hosting story attached. The awkwardness is internal: the open Large release beside it is close enough on quality that the closed tier has to justify itself on features rather than results.
Where it shines
- Strong tool calling, which is what agent frameworks actually stress
- Materially cheaper than American frontier models
- EU hosting without a bespoke contract
Worth knowing first
- Closed weights, unlike the Large release beside it
- Neither the cheapest nor the strongest option in its own line-up
Index score
78.0
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 50 of 73
- Builds
- LLMs
- Category
- Frontier
- Output
- Text, code and images in, text out
- Pricing
- Paid only
- Best for
- Agentic and multimodal work with EU hosting
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 01Claude Opus 5.5Claude Opus 5.5ProvisionalFable-level answers for less than the Opus it replaces. The new default.Frontier modelPaid only92.8
- 03Gemini 4 ArgonGemini 4 ArgonProvisionalTop of the human-preference leaderboard, and almost nobody can use it yet.Frontier modelPaid only90.6
- 04GPT-6.1 SolGPT-6.1 SolProvisionalNear-Astra answers at a fifth of the price. The OpenAI model most teams should deploy now.Frontier modelPaid only90.3
