Gemini Flash vs Hunyuan OCR
Pick Gemini Flash for cheap bulk PDF to Markdown or JSON extraction through a hosted API, especially with mixed-language documents. Pick Hunyuan OCR when you must self-host, need bounding boxes, or work on photo text, image translation or historical scripts, and you are outside the EU, UK and South Korea. The main split is hosted per-token API versus a free, self-hosted 1B model with a restrictive licence.
Gemini Flashgoogle/gemini-flash | Hunyuan OCRtencent/hunyuan-ocr | |
|---|---|---|
| Type | Vision model | Open source |
| Price | 15 credits / page | 2 credits / page |
| Per 1,000 pages | $15.00 | $2.00 |
| Typical speed | 1.6 s / page | 1.5 s / page |
| Best for | Handwriting and mixed layouts | Chinese documents and photos |
| Inputs | PDF, images | PDF, images |
| Languages | 100+ languages | 100+ languages |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | No | ✓ Yes |
| 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.
Which to choose
Choose Gemini Flash if
- Gemini 3 Flash scores 92.62 on OmniDocBench v1.6, close to Gemini 3 Pro at 92.91.
- You call an API with no GPU or CUDA setup, and the 1M-token context window fits long documents in one request.
- It suits cheap bulk PDF to Markdown, prompt or schema-driven JSON extraction, and mixed-language documents, at $0.50 input and $3.00 output per 1M tokens ($0.25 and $1.50 with Batch or Flex).
Choose Hunyuan OCR if
- You can run it yourself at no licence fee, even on a laptop through llama.cpp GGUF, and you only pay for compute.
- One model of about 1B parameters covers document parsing, text spotting, information extraction and image translation, and it returns bounding boxes.
- It handles photo and scene text, low-resource and ancient scripts, with 12 task types exposed through the vLLM client.
Key differences
| Gemini Flash | Hunyuan OCR | |
|---|---|---|
| How you run it | Hosted API (Gemini API or Vertex AI) | Self-host via vLLM, Transformers or llama.cpp |
| Licence | Proprietary | Tencent Hunyuan Community License; excludes EU, UK and South Korea; separate licence above 100M monthly active users |
| Price | Gemini 3 Flash Preview: $0.50 input and $3.00 output per 1M tokens; Batch/Flex $0.25 and $1.50 | Free software; you pay for compute |
| Output formats | Markdown, JSON, HTML, plain text | Markdown, HTML, LaTeX, JSON, plain text, bounding boxes |
| Benchmark | OmniDocBench v1.6 overall: 92.62 (Gemini 3 Flash) | ParseBench-2 overall: 67.11 (v1.5); ParseBench mean: 63.86. Not published on OmniDocBench |
| Hardware and setup | None on your side; air-gapped or on-prem deployment is not possible | NVIDIA GPU with CUDA 13 recommended for the unified vLLM setup; llama.cpp GGUF path for CPU, consumer GPU or laptop |
| Known weakness | Cost per page varies with page density and output length; output is prompt-driven with no fixed schema or confidence scores; several newer Flash models rise from $0.75/$3.75 to $1.50/$7.50 per 1M tokens on January 1, 2027 | An archived 1.0 note said Transformers and vLLM outputs differed in accuracy; text score 62.88 on ParseBench-2 in public CI replay |
Sources
- Gemini API pricing
- OmniDocBench leaderboard
- Gemini pricing changes 2026
- HunyuanOCR GitHub
- HunyuanOCR model card
- Hunyuan Community License
What each one is
Gemini Flash
General vision model with strong handwriting and layout understanding at a low price.
Gemini Flash 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": "google/gemini-flash","fallbacks": ["tencent/hunyuan-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Gemini Flash alternativesHunyuan OCR alternativesBest OCR models in 2026