Model reference · open weights
DeepSeek-OCR is an open-weight language model from deepseek-ai, listed in the AxForge catalogue. AxForge can bring it up on EU-owned hardware for you on request — with the licence handled where one is required.
About
Usage Inference using Huggingface transformers on NVIDIA GPUs. Requirements tested on python 3.12.9 + CUDA11.8： vLLM Refer to 🌟GitHub for guidance on model inference acceleration and PDF processing, etc.<!-- -- [2025/10/23] 🚀🚀🚀 DeepSeek-OCR is now officially supported in upstream vLLM. Visualizations Acknowledgement We would like to thank Vary, GOT-OCR2.0, MinerU, PaddleOCR, OneChart, Slow Perception for their valuable models and ideas. We also appreciate the benchmarks: Fox, OminiDocBench. Citation
Summarised from the published model card. Read the full card on the HuggingFace links below.
Specifications
| Maker | deepseek-ai |
|---|---|
| Type | Language models |
| Parameters (lead) | 3.3B |
| Context | 8k tokens |
| Variants | 1 |
| Runs with | transformers |
| Released | 2025-10-17 |
| Popularity | 2.4M downloads / month |
| Likes | 3,348 |
| Licence | Open weights |
How it works
Variants
Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.
| Variant | Params | Precision | VRAM | Fits 16 GB | Weights |
|---|---|---|---|---|---|
| DeepSeek-OCR | 3.3B | BF16 | ~7.7 GB | ✓ | Weights ↗ |
Using it via the API
Once AxForge deploys deepseek-ocr for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (deepseek-ocr below is illustrative; you get the exact model name on deployment.)
$ curl -sS https://api.axforge.ai/v1/chat/completions \
-H "Authorization: Bearer $AXFORGE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"deepseek-ocr","messages":[{"role":"user","content":"Hello"}]}'
Licence
Open weights under mit — commercial use is permitted. Deploy it on AxForge EU hardware on request. Read the licence ↗