Mistral Large
Mistral's premium family for multilingual reasoning, coding, and enterprise copilots.
- Creator
- Mistral AI
- Lifecycle
- Active
- Context
- 128.0K
- Max output
- 8.2K
- Released
- Feb 26, 2024
- Status
- up
- Input
- $2.00 / 1M tokens
- Output
- $6.00 / 1M tokens
- Cached read
- — / 1M tokens
- Cached write
- — / 1M tokens
- Batch discount
- 50%
- Source
- Mistral pricing
- Verified
- Apr 1, 2026 (Medium)
Capabilities
- Modalities
- textimage→text
- Capabilities
- reasoningimageInputbatchSupportfunctionCallingstructuredOutputs
- Strengths
- Competitive price-performance, Strong multilingual handling
- Tradeoffs
- Smaller host ecosystem than OpenAI or Anthropic
Official Links
Benchmark Coverage
| Benchmark | Version | Score | Date | Source | Notes |
|---|---|---|---|---|---|
| MMLU | 2025-rolling | 81.6 accuracy | Nov 14, 2025 | Mistral docs | Use as raw evidence only. |
Release History
| Release | Alias | Lifecycle | Release Date | Deprecation | Shutdown | Summary |
|---|---|---|---|---|---|---|
| Mistral Large | mistral-large | Active | Feb 26, 2024 | — | — | Current premium Mistral family. |
Host Coverage
| Host | Type | Context | Pricing Note | Differences |
|---|---|---|---|---|
| Mistral API | first-party | 128.0K | Reference pricing and official aliases. | Direct vendor access |
| AWS Bedrock | cloud | 128.0K | Managed cloud hosting for teams standardizing on AWS. | AWS IAM integration |
Migration Guidance
Useful price-performance step up from small open-weight families when multilingual reasoning and stronger tool use matter.
Breaking changes: Prompt tuning may need adjustment versus proprietary-provider instruction formats.
Change Events
| Date | Type | Title | Description | Source |
|---|---|---|---|---|
| Feb 26, 2024 | family_added | Family launched | Mistral Large launched as Mistral's premium production family. | Mistral changelog |
Other models from Mistral AI
Codestral, Devstral 2, Mistral Large 3, Mistral Medium 3.1, Mistral Small 4