OCRRouter

Unlimited OCR vs Qwen-VL

Pick Unlimited OCR to turn long reports, books and bulk PDFs into Markdown with bounding boxes for RAG. It parses 40+ pages in one pass under an MIT licence. Pick Qwen-VL when OCR is one task among several, such as VQA, photos or screenshots with text, or JSON output, and you want one general model.

Unlimited OCRbaidu/unlimited-ocrQwen-VLqwen/qwen-vl
TypeOpen sourceOpen source
Price2 credits / page2 credits / page
Per 1,000 pages$2.00$2.00
Typical speed1.2 s / page2.0 s / page
Best forBooks and very long documentsFlexible extraction with prompts
InputsPDF, imagesPDF, images
Languages30+ languages30+ languages
Max pages per request1,000200
Tables✓ Yes✓ Yes
HandwritingNo✓ Yes
Math✓ Yes✓ Yes
Bounding boxes✓ Yes✓ Yes
ConfidenceNoNo
FormsNoNo
Multilingual✓ Yes✓ Yes
On OCRRouterAvailableAvailable

Prices are what you pay on OCRRouter. Speed is the provider's typical figure, not a benchmark we ran.

Which to choose

Choose Unlimited OCR if

  • It parses 40+ pages in a single forward pass with roughly constant memory, so long reports and books go in as one request.
  • Unlimited OCR 2 scores 77.1 on MarathonDocBench versus 33.7 for the next model, per Baidu.
  • It is MIT licensed and emits Markdown, plain text, block types and bounding boxes.

Choose Qwen-VL if

  • It is a general vision-language model, so one deployment covers OCR, reasoning and VQA on photos and screenshots.
  • It reads 32 languages and handles low light, blur and tilted text, per Qwen.
  • It comes in sizes from 2B to 235B-A22B under Apache-2.0 and can output JSON and HTML as well as Markdown and boxes.

Key differences

Unlimited OCRQwen-VL
Main jobMulti-page document parser, 40+ pages in one passGeneral vision-language model with OCR and document parsing
LicenceMITApache-2.0 (Qwen3-VL repository)
Output formatsMarkdown, plain text, bounding boxesPlain text, HTML, bounding boxes, JSON, Markdown
Model size3B total, about 500M active per token (Unlimited OCR 2 is 3.34B)2B, 4B, 8B, 32B dense; 30B-A3B and 235B-A22B MoE
HardwareNVIDIA GPU with CUDA 12.9 or 13.0; version 2 under SGLang runs out of memory on 12 GB GPUsFP8 checkpoints need an H100-class GPU and CUDA 12+; the 235B example uses 8-way tensor parallel on vLLM
BenchmarkParseBench mean 46.17 (independent); MarathonDocBench 77.1 for version 2 (authors)olmOCR-bench 64.6 for Qwen 3 VL 8B (independent), below dedicated OCR models
Known failureRepetition and looping on some pages, hallucinated text, bbox coordinate loops on SGLang with GB10 GPUs, and most text styling lost (formatting score 0.97 on ParseBench)Repeating loops, worse with greedy decoding and on low-quality images; one transformers version caused loops on 10-37% of OCR outputs versus 2-3% baseline

Sources

What each one is

Unlimited OCR

Open-source model from Baidu that reads hundreds of pages in one pass with flat memory use.

Unlimited OCR details

Qwen-VL

Alibaba's open-weight vision language models (Qwen2.5-VL, Qwen3-VL). Strong general OCR and grounding.

Qwen-VL details

Try both on your own file

Run the file once with each model and compare the output.

PDF, PNG, JPG, WebP, HEIC. Up to 20 MB each, 10 files.No file?

Use one, fall back to the other

{"model": "baidu/unlimited-ocr","fallbacks": ["qwen/qwen-vl"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

Unlimited OCR alternativesQwen-VL alternativesBest OCR models in 2026