Scanned document extraction
What AI can do for this task
Extract text and data from scanned pages or photos of documents, sent to the model as images. Vision support is required, but it does not prove OCR accuracy: that would need measurement. For scanned PDF files sent as files, see PDF document extraction.
Technical requirements
- Image input (capability listed for the offering)
Scanned pages are images, so the model must accept images. Declared vision support does not demonstrate OCR accuracy.
AI offerings and technical compatibility
Of 20 offerings in the catalog: 18 compatible, 2 unknown, 0 incompatible.
- Compatible
- the technical requirements are met according to the available information.
- Unknown
- the available information is not enough to decide. This is not the same as incompatible.
- Incompatible
- at least one requirement is known not to be met.
Within each group, offerings are listed alphabetically by producer and model. The order says nothing about quality.
Compatible 18
| Model | Version | Offering (access) | Requirements | Source |
|---|---|---|---|---|
| Claude Haiku 4.5 Anthropic |
claude-haiku-4-5 | OpenRouteranthropic/claude-haiku-4.5 |
|
OpenRouter API 2026-10-06 |
| Claude Opus 4.8 Anthropic |
claude-opus-4-8 | OpenRouteranthropic/claude-opus-4.8 |
|
OpenRouter API 2026-10-06 |
| Claude Opus 5.5 Anthropic |
claude-opus-5-5 | OpenRouteranthropic/claude-opus-5.5 |
|
OpenRouter API 2026-10-06 |
| Claude Sonnet 5.5 Anthropic |
claude-sonnet-5-5 | OpenRouteranthropic/claude-sonnet-5.5 |
|
OpenRouter API 2026-10-06 |
| DeepSeek V4.1 Flash DeepSeek |
deepseek-v4.1-flash | OpenRouterdeepseek/deepseek-v4.1-flash |
|
OpenRouter API 2026-10-06 |
| Gemini 3.1 Pro (Preview) |
gemini-3.1-pro-preview | OpenRoutergoogle/gemini-3.1-pro-preview |
|
OpenRouter API 2026-10-06 |
| Gemini 3.8 Flash |
gemini-3.8-flash | OpenRoutergoogle/gemini-3.8-flash |
|
OpenRouter API 2026-10-06 |
| Llama 4 Maverick Meta |
Llama-4-Maverick-17B-128E-Instruct | OpenRoutermeta-llama/llama-4-maverick |
|
OpenRouter API 2026-10-06 |
| Llama 4 Scout Meta |
Llama-4-Scout-17B-16E-Instruct | OpenRoutermeta-llama/llama-4-scout |
|
OpenRouter API 2026-10-06 |
| Ministral 3 8B Mistral AI |
v25.12 | OpenRoutermistralai/ministral-8b-2512 |
|
OpenRouter API 2026-10-06 |
| Mistral Large 3 Mistral AI |
v25.12 | OpenRoutermistralai/mistral-large-2512 |
|
OpenRouter API 2026-10-06 |
| Mistral Medium 3.5 Mistral AI |
v26.04 | OpenRoutermistralai/mistral-medium-3-5 |
|
OpenRouter API 2026-10-06 |
| Mistral Small 4 Mistral AI |
v26.03 | OpenRoutermistralai/mistral-small-2603 |
|
OpenRouter API 2026-10-06 |
| GPT-6 Astra OpenAI |
gpt-6-astra | OpenRouteropenai/gpt-6-astra |
|
OpenRouter API 2026-10-06 |
| GPT-6 Luna OpenAI |
gpt-6-luna | OpenRouteropenai/gpt-6-luna |
|
OpenRouter API 2026-10-06 |
| GPT-6.1 Sol OpenAI |
gpt-6.1-sol | OpenRouteropenai/gpt-6.1-sol |
|
OpenRouter API 2026-10-06 |
| Qwen3.8 27B Qwen (Alibaba) |
Qwen3.8-27B | OpenRouterqwen/qwen3.8-27b |
|
OpenRouter API 2026-10-06 |
| Grok 4.7 xAI |
grok-4.7 | OpenRouterx-ai/grok-4.7 |
|
OpenRouter API 2026-10-06 |
Unknown 2
| Model | Version | Offering (access) | Requirements | Source |
|---|---|---|---|---|
| DeepSeek V4 Pro DeepSeek |
deepseek-v4-pro | OpenRouterdeepseek/deepseek-v4-pro-0813 |
|
— |
| Qwen3.8 2.4T-A95B Qwen (Alibaba) |
Qwen3.8-2.4T-A95B | OpenRouterqwen/qwen3.8-2.4t-a95b |
|
— |
How much can you trust this comparison?
Technical compatibility is not a measure of answer quality.
- 18 capability checks on this page rely only on the catalog source recorded for each offering (shown in the Source column), not on independent measurement.
- 2 capability checks are not listed in the catalog for their offerings: they are shown as unknown, not as missing.
- No capability on this page has been measured independently by iaright yet: listed capabilities are not verified performance.
- Capabilities and limits are based on the source identified for each offering, not necessarily on the model maker's documentation. OpenRouter catalog data may differ from model-maker documentation or from the limits of a specific upstream provider.
- Vision support listed for an offering does not demonstrate OCR accuracy or correct reading of images.
- iaright has not yet published comparative measurements of quality, cost, speed or reliability for this task.