Mistral OCR Models by MistralAI
MistralAIProprietary
2 models2025
Last refreshed 2026-05-19. Next refresh: weekly.
Details
ResearcherMistralAI
LicenseProprietary
Commercial useCommercial use: conditional
Models2
Released2025
Capabilities
VisionAll models
MultimodalAll models
Links
WebsiteAbout
Mistral's document AI and optical character recognition model family, designed for extracting structured information from documents, PDFs, and images.
Decision facts
- Best fit
- document aivision and multimodal workstructured outputs
- Capability starting point
- Mistral Document AI 2512 with multimodal inputs
- Lowest tracked input
- Not tracked
- Closest related family
- Devstral
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
2 in view
Mistral Document AI 2512Current
Use when the workload needs document ai and multimodal inputs.
2025-12document aimultimodal inputs
Mistral Document AI 2505Current
Use when the workload needs multimodal inputs.
2025-05multimodal inputs
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Mistral Document AI 2512 | Use when the workload needs document ai and multimodal inputs. | 2025-12 | document aimultimodal inputs | Current |
| Mistral Document AI 2505 | Use when the workload needs multimodal inputs. | 2025-05 | multimodal inputs | Current |
Release Timeline
2 release groups2025-12
1 current
Mistral Document AI 2512
Currentdocument aimultimodal inputs
2025-05
1 current
Mistral Document AI 2505
Currentmultimodal inputs
Specifications(2 models)
| Model | Released | Vision | Multimodal |
|---|---|---|---|
| Mistral Document AI 2512 | 2025-12 | Yes | Yes |
| Mistral Document AI 2505 | 2025-05 | Yes | Yes |
