Chandra vs Hunyuan OCR
Pick Chandra for multilingual scanned documents, PDF to Markdown for RAG, and pages with tables, forms and math, especially if you want a hosted API or can stay under the $2M funding/revenue licence line. Pick Hunyuan OCR if you need a ~1B model that runs on a laptop or consumer GPU, or need text spotting, image translation and low-resource scripts, and you are outside the EU, UK and South Korea.
Chandradatalab/chandra | Hunyuan OCRtencent/hunyuan-ocr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 2 credits / page | 2 credits / page |
| Per 1,000 pages | $2.00 | $2.00 |
| Typical speed | 2.0 s / page | 1.5 s / page |
| Best for | Complex layouts and handwriting | Chinese documents and photos |
| Inputs | PDF, images | PDF, images |
| Languages | 40+ languages | 100+ languages |
| Max pages per request | 500 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | No | No |
| Forms | ✓ Yes | 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.
Which to choose
Choose Chandra if
- Chandra 2 scores 85.8 on olmOCR-bench and averages 77.8% across 43 languages, so it suits multilingual scanned documents.
- You can skip running GPUs by using the hosted Datalab API at $1, $3 or $7.50 per 1,000 pages, with a free tier of $20/month.
- It returns Markdown, HTML and JSON with bounding boxes and labels per block, which fits PDF to Markdown pipelines for RAG.
Choose Hunyuan OCR if
- At about 1B parameters it runs on laptops, CPU or consumer GPUs through the llama.cpp GGUF path.
- One model covers document parsing, text spotting, information extraction and image translation, with 12 task types in the vLLM client.
- It targets low-resource and ancient-script OCR, where Chandra is weak (43.2% Urdu, 51.1% Sanskrit).
Key differences
| Chandra | Hunyuan OCR | |
|---|---|---|
| How you run it | Self-host with vLLM or Hugging Face, or use the hosted Datalab API | Self-host only, via vLLM, Transformers or llama.cpp |
| Licence | Code Apache-2.0; weights modified OpenRAIL-M, free under $2M funding/revenue, paid licence above that | Tencent Hunyuan Community License; excludes EU, UK and South Korea; separate licence above 100M monthly active users |
| Price | Self-hosted free within licence terms; API $1, $3 or $7.50 per 1,000 pages | Free, self-hosted; you pay for compute |
| Hardware | Single NVIDIA GPU via vLLM; presets for A100, L40S and RTX 4090/5090 | NVIDIA GPU with CUDA 13 for vLLM; GGUF path for CPU, consumer GPU or laptop |
| Output formats | Markdown, HTML, JSON, bounding boxes | Markdown, HTML, LaTeX, JSON, plain text, bounding boxes |
| Benchmark | olmOCR-bench overall 85.8 (Chandra 2, authors' table) | ParseBench-2 overall 67.11 (v1.5, independent) |
| Known failure | Token-loop hallucinations on some Russian/Kazakh scans; Old Scans only 51.1 on olmOCR-bench | Transformers and vLLM outputs differed in accuracy in the archived 1.0 note; needs CUDA 13 and vLLM 0.25.1+ |
Sources
- Chandra GitHub README
- Datalab pricing
- GitHub issue #51 token loops
- HunyuanOCR model card
- Hunyuan Community License
What each one is
Chandra
Datalab's OCR model that keeps layout and handles forms, tables and handwriting.
Chandra detailsHunyuan OCR
Small end-to-end OCR model from Tencent covering parsing, extraction and photo translation.
Hunyuan OCR 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/chandra","fallbacks": ["tencent/hunyuan-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Chandra alternativesHunyuan OCR alternativesBest OCR models in 2026