Mistral OCR vs MiniCPM-V 4.5
Pick Mistral OCR if you want a hosted API that turns scanned business PDFs into Markdown and HTML tables at $2 per 1,000 pages, with no GPUs to run. Pick MiniCPM-V 4.5 if the documents must stay on your own hardware or on-device, since it is Apache-2.0 and you self-host it. The deciding factor is deployment: a priced API against open weights you run yourself.
Mistral OCRmistral/mistral-ocr | MiniCPM-V 4.5openbmb/minicpm-v | |
|---|---|---|
| Type | OCR API | Vision model |
| Price | 25 credits / page | 2 credits / page |
| Per 1,000 pages | $25.00 | $2.00 |
| Typical speed | 0.9 s / page | 2.5 s / page |
| Best for | High-volume PDF to Markdown | OCR on edge devices |
| Inputs | PDF, images | images |
| Languages | 100+ languages | 30+ languages |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | ✓ Yes |
| Math | ✓ Yes | No |
| Bounding boxes | ✓ Yes | No |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | ✓ Yes | ✓ Yes |
| On OCRRouter | Available | Available |
Prices are what you pay on OCRRouter. Speed is the provider's typical figure, not a benchmark we ran.
Which to choose
Choose Mistral OCR if
- It is a hosted API at $2 per 1,000 pages, or $1 per 1,000 through the Batch API, so high-volume PDF to Markdown for RAG needs no infrastructure.
- OCR 3 rebuilds tables as HTML with rowspan and colspan, and it also returns JSON and extracted images.
- It has independent OmniDocBench scores (78.83 on v1.5 for the first version, 85.66 on v1.6), which you can check on the public leaderboards.
Choose MiniCPM-V 4.5 if
- It is Apache-2.0 with open weights and code, so documents never leave your machines and there is no per-page fee, only compute cost.
- It runs on llama.cpp and vLLM with int4, GGUF and AWQ builds, which suits on-device or edge OCR and local document Q&A on a single GPU.
- At 8.7B parameters it is a general vision language model that can answer questions about a document, and it outputs plain text, Markdown or JSON.
Key differences
| Mistral OCR | MiniCPM-V 4.5 | |
|---|---|---|
| How you run it | Hosted API (current pinned version mistral-ocr-2512, OCR 3) | Self-host with llama.cpp, vLLM, Ollama or Transformers; Ollama and SGLang were still pending official releases on the model card |
| Licence | Proprietary | Apache-2.0 weights and code |
| Price | $2 per 1,000 pages; $3 with structured annotations; $1 per 1,000 via Batch API | Free to use; you pay for compute |
| Output formats | Markdown, HTML, JSON, images | Plain text, Markdown, JSON; no bounding box or layout output |
| Benchmarks | OmniDocBench v1.5 overall 78.83 (Mistral OCR 2503); v1.6 overall 85.66 | OpenBMB claims leading OCRBench results, but the model card gives scores only as images, so the numbers are hard to verify |
| Tables | OCR 3 reconstructs tables as HTML with rowspan and colspan | Not published |
| Known failure modes | Hallucinated text on unreadable handwriting; colored-background pages returned as images with no OCR text (2025 release); changed checkbox states and misread characters on a dense financial table | No int4 VRAM figure published; no coordinates, so it does not fit jobs that need them |
Sources
- Mistral OCR 3 announcement
- VentureBeat on OCR 3 pricing
- OmniDocBench v1.6 leaderboard
- Reducto test of Mistral OCR
- Hugging Face model card
- GitHub repo
What each one is
Mistral OCR
Fast document OCR that returns clean Markdown with tables and embedded images.
Mistral OCR detailsMiniCPM-V 4.5
8B vision model from OpenBMB (August 2025) built to run on devices, with strong OCR and document parsing.
MiniCPM-V 4.5 detailsTry both on your own file
Run the file once with each model and compare the output.
Use one, fall back to the other
{"model": "mistral/mistral-ocr","fallbacks": ["openbmb/minicpm-v"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Mistral OCR alternativesMiniCPM-V 4.5 alternativesBest OCR models in 2026