All AI models · LLM benchmarks · Methodology

DeepSeek

deepseek-ai/DeepSeek-OCR

OCR model for extracting structured text from documents and screenshots

Model facts

Context window
8K tokens
Maximum output
8K tokens
Cheapest paid input
free per 1M tokens
Cheapest paid output
free per 1M tokens
Open weights
yes
Providers
1

Capabilities and modalities

attachments, temperature control, open weights, text, image.

Local hardware estimate

At 4K context: Q4_K_M 3.0 GiB, Q8_0 4.6 GiB, F16 7.5 GiB working memory.

Parameters
3.3 billion
Architecture
deepseekocr
Model license
mit

Planning estimate adapted from the MIT-licensed whichllm estimator using metadata from Hugging Face; not a vendor minimum requirement.

Observed price history

  • 2026-08-24: free input / free output per 1M tokens

API providers

  • SiliconFlow (China) — model id deepseek-ai/DeepSeek-OCR; free input / free output per 1M tokens; provider documentation