Marker vs DeepSeek-VL2
Pick Marker to convert whole PDFs, such as books and papers with equations, into Markdown, HTML or JSON on your own hardware, and to chunk them for RAG. Pick DeepSeek-VL2 to answer questions about a single page or chart, where it scores 93.3 on DocVQA, and when its weights licence allows commercial use. The main split is the job: Marker is a PDF converter, DeepSeek-VL2 is a vision language model.
Markerdatalab/marker | DeepSeek-VL2deepseek/deepseek-vl2 | |
|---|---|---|
| Type | OCR API | Vision model |
| Price | 14 credits / page | 2 credits / page |
| Per 1,000 pages | $14.00 | $2.00 |
| Typical speed | 2.5 s / page | 3.0 s / page |
| Best for | Books and research papers | Charts and tables |
| Inputs | PDF, images, DOCX | images |
| Languages | 90+ languages | Mainly English and Chinese |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| Math | ✓ Yes | No |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | ✓ Yes | No |
| 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 Marker if
- You want PDF to Markdown, HTML, JSON block trees or RAG chunks, plus LaTeX output for equations.
- You have no big GPU: fast mode runs on CPU at 23.6 pages per second with about 18 MB of ONNX weights, though it drops to 66.6 on olmOCR-bench.
- You want a hosted option too, since Datalab's API includes about 5,000 free pages per month.
Choose DeepSeek-VL2 if
- You need to answer questions about documents and charts, with DocVQA 93.3 and ChartQA 86.0 for the full model.
- Your company is above $5M and needs commercial use without a paid weight licence, since the DeepSeek Model License allows it.
- You work one page at a time and can give it a GPU: Tiny runs under 40GB, and Small scores 834 on OCRBench.
Key differences
| Marker | DeepSeek-VL2 | |
|---|---|---|
| What it is | PDF to Markdown converter built on the Surya OCR models | Mixture-of-experts vision language model in Tiny, Small and full sizes |
| Licence | Code Apache-2.0; weights under a modified AI Pubs Open Rail-M licence, free for research, personal use and startups under $5M | MIT for code; DeepSeek Model License for weights, commercial use allowed |
| How you run it | Self-host on GPU, CPU or Apple Silicon, or use Datalab's hosted API | Self-host the open weights only |
| Hardware | CPU fast mode with about 18 MB of ONNX weights; vllm mode needs an NVIDIA GPU | Small may need 80GB GPU memory, or 40GB with incremental prefilling; Tiny runs under 40GB |
| Output formats | Markdown, HTML, JSON, chunks, LaTeX, images | Plain text, Markdown |
| Benchmarks | olmOCR-bench 83.7 and OmniParseBench 85.9 in vllm mode, both reported by Datalab | DocVQA 93.3, OCRBench 834 (Small), ChartQA 86.0, all from the authors' paper |
| Known failure | CPU fast mode drops to 66.6 on olmOCR-bench | Struggles with blurry images; supports only a few images per session, and dynamic tiling turns off with more than two images |
Sources
What each one is
Marker
Hosted version of the open-source Marker and Surya models. Strong on books and papers.
Marker detailsDeepSeek-VL2
DeepSeek's mixture-of-experts vision models (December 2024) with OCR and table and chart understanding.
DeepSeek-VL2 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": "datalab/marker","fallbacks": ["deepseek/deepseek-vl2"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Marker alternativesDeepSeek-VL2 alternativesBest OCR models in 2026