DeepSeek OCR 2 is an open-source language model by deepseek-ai. Features: 3.4b LLM, VRAM: 6.8GB, Context: 8K, License: apache-2.0, LLM Explorer Score: 0.49.
Arxiv:2510.18234 Arxiv:2601.20552 Custom code Deepseek Deepseek vl v2 Eval-results Feature-extraction Image-text-to-text Multilingual Ocr Region:us Safetensors Sharded Tensorflow Vision Vision-language
Model Card on HF 🤗: https://huggingface.co/deepseek-ai/DeepSeek-OCR-2
DeepSeek OCR 2 Benchmarks
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
DeepSeek OCR 2 Parameters and Internals
| LLM Name | DeepSeek OCR 2 |
| Repository 🤗 | https://huggingface.co/deepseek-ai/DeepSeek-OCR-2 |
| Model Size | 3.4b |
| Required VRAM | 6.8 GB |
| Updated | 2026-05-20 |
| Maintainer | deepseek-ai |
| Model Type | deepseek_vl_v2 |
| Model Files | |
| Model Architecture | DeepseekOCR2ForCausalLM |
| License | apache-2.0 |
| Context Length | 8192 |
| Model Max Length | 8192 |
| Transformers Version | 4.46.3 |
| Tokenizer Class | LlamaTokenizerFast |
| Padding Token | <｜▁pad▁｜> |
| Vocabulary Size | 129280 |
| Torch Data Type | bfloat16 |
Rank the DeepSeek OCR 2 Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 53813 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a