Gemma 3google/gemma-3 | Aya Visioncohere/aya-vision | |
|---|---|---|
| Type | Open source | Open source |
| Price | 2 credits / page | 1 credit / page |
| Per 1,000 pages | $2.00 | $1.00 |
| Typical speed | 3.0 s / page | 3.0 s / page |
| Best for | Self-hosted image reading | Multilingual image questions |
| Inputs | images | images |
| Languages | 140+ languages | 23 languages |
| Max pages per request | 50 | 50 |
| Tables | No | No |
| Handwriting | No | No |
| Math | No | No |
| Bounding boxes | No | 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.
What each one is
Gemma 3
Google's open-weight model family (March 2025). The 4B, 12B and 27B sizes accept images.
Gemma 3 detailsAya Vision
Cohere's open-weight multilingual vision models (8B and 32B, March 2025).
Aya Vision detailsTry both on your own file
Run the file once with each model and compare the output.
PDF, PNG, JPG, WebP, HEIC. Up to 10 files at onceNo file?
Use one, fall back to the other
{"model": "google/gemma-3","fallbacks": ["cohere/aya-vision"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Gemma 3 vs Command A VisionGemma 3 vs LLaVA-OneVisionQwen3-VL vs Aya VisionGemini 2.0 Flash vs Aya Vision
Gemma 3 alternativesAya Vision alternativesBest OCR models in 2026