Pixtral OCR online
Mistral's open-weight vision models, Pixtral 12B and Pixtral Large. Mistral's API no longer serves them, so they run self-hosted.
PDF, PNG, JPG, WebP, HEIC. Up to 20 MB each, 10 files.No file?
- Model ID
- mistral/pixtral
- Type
- Vision model
- Price
- 2 credits / page ($2.00 per 1,000 pages)
- Typical speed
- 3.0 s / page
- Inputs
- images
- Languages
- Multilingual
- Max pages per request
- 500
- Output
- Markdown, JSON
- License
- Pixtral 12B: Apache-2.0. Pixtral Large: Mistral Research License
- Supports
- Multilingual
curl https://api.ocrrouter.ai/v1/ocr \
-H "Authorization: Bearer $OCRROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "mistral/pixtral",
"file_url": "https://example.com/invoice.pdf",
"format": "markdown"
}'Pixtral speed and reliability
Loading…
Measured on our own traffic over the last 7 days.
Pixtral pricing
Pixtral costs 2 credits per page. One dollar buys 1,000 credits, so 1,000 pages cost $2.00. You only pay for pages that return a result.
New accounts get 50 free credits, enough for about 25 pages with this model. See pricing for credit packs.
Similar models
1Amazon Nova LiteLow-cost Amazon Nova model with image and document input.
amazon/nova-lite1 credit / page2Reka FlashReka's multimodal model with image, video and audio input.reka/reka-flash3 credits / page3Grok 2 VisionxAI's earlier vision model for image understanding.xai/grok-2-vision6 credits / pageAll Pixtral alternativesBest OCR models by use case
Compare Pixtral
Pixtral vs Amazon Nova LitePixtral vs Reka FlashPixtral vs Grok 2 VisionPixtral vs Typhoon OCRPixtral vs MolmoPixtral vs o3