flagship tier · 2024
Mistral AI
Mistral Large 2
French-built flagship with strong code, reasoning, and multilingual capability.
Context window
128k tokens
Peak context for this model.
Availability
Mistral API (La Plateforme), Azure AI, self-hosted via deployment partners
Where you can run it.
Modalities
Text · Code
Input/output coverage.
Pricing
$2.00 / 1M input tokens, $6.00 / 1M output tokens
Latency: Low to medium depending on provider
Strengths
- Competitive reasoning with concise answers and good tool-use adherence.
- Multilingual strength across European languages.
- Great price-performance for production workloads.
Best for
- APIs that need predictable cost while keeping quality high.
- Developer tooling, code review, and refactors.
- Customer experience workloads across languages.
Summary
- Tier: flagship
- Release: 2024
- Latency: Low to medium depending on provider