MODELS
Mistral Large 2.5
Mistral's flagship text model with strong tool use and European hosting options.
Specs
- Context window
- 128,000
- Max output
- 16,384
- Modalities
- text
- Tool use
- ✓
- Vision
- —
- Streaming
- ✓
- License
- proprietary
- Released
- 2025-07-01
Pricing
- Input / 1M
- $2.00
- Output / 1M
- $6.00
Cost estimate
Mistral Large 2.5 is the company's top-tier proprietary text model, aimed at agent workflows and multilingual business apps. It handles 128K context, supports streaming and function calling, and prices at $2/$6 per million tokens — between GPT-4o mini and the full GPT-4o. No vision, no audio: it's a pure text and tool-use model. European builders pick it largely for data-residency reasons and for its consistently strong French, German, and Spanish output.
Editor's verdict
Pick it if you need EU hosting, GDPR-friendly contracts, or genuinely good non-English European languages — that's its real moat. On pure capability it trails Claude Sonnet 4 and GPT-4o, especially on long reasoning chains and code. The lack of vision is also a real limit in 2025. Solid default for European agent stacks; harder to justify elsewhere when peers cost the same or less.
Reviews
No reviews yet. Be the first.
Last updated: 2026-04-29