DeepSeek-OCR
DeepSeek's 3B MIT-licensed OCR specialist built on DeepEncoder, notable for its 'optical compression' approach. Punches well above its weight on documents, LaTeX, and tables.
By DeepSeek · China
Updated 2026-09-15
When to pick this model
- High-volume document OCR pipelines
- Extracting LaTeX formulas from scientific papers
- Parsing tables from PDFs, scans, and receipts
- Edge deployments needing OCR in ~2 GB VRAM
- MIT-licensed alternative to closed OCR APIs
VRAM requirements by quantization
| Quantization | VRAM required |
|---|---|
| Q4_K_M (recommended) | 2 GB |
| Q5_K_M | 2.5 GB |
| Q8_0 | 4 GB |
| FP16 (no quantization) | 6 GB |
VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.
In practice, DeepSeek-OCR fits an 8 GB consumer card at Q4_K_M (2 GB). Stepping up to Q8_0 nearly doubles the footprint to 4 GB, and unquantized FP16 weights take 6 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.
Without a GPU, DeepSeek-OCR needs roughly 5 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 25 tokens/sec on entry-level GPUs, on the order of 70 tokens/sec on a mid-range card, and up to 160 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.
What hardware do you need
The table below matches DeepSeek-OCR to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.
| GPU memory | Example cards | Best fit for DeepSeek-OCR |
|---|---|---|
| 8 GB | RTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GB | FP16 (6 GB used) |
| 12 GB | RTX 5070, RTX 5070 Ti Laptop, RTX 4080 Laptop | FP16 (6 GB used) |
| 16 GB | RTX 5080, RTX 4080 Super, Radeon RX 9070 XT | FP16 (6 GB used) |
| 24 GB | RTX 4090, Radeon RX 7900 XTX, RTX 5090 Laptop | FP16 (6 GB used) |
| 32 GB | RTX 5090 | FP16 (6 GB used) |
Which hardware should you buy to run DeepSeek-OCR?
To run DeepSeek-OCR locally at Q4, you need ~2 GB for Q4 weights alone. Hardware option to compare: RTX 5060 Ti 16GB (ASUS Prime). Leave memory for the system and context; verify inference-engine support. A mini PC does not provide CUDA or macOS/MLX.
As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.
Strengths
- Best-in-class OCR quality at only 3B parameters
- Handles LaTeX formulas and table structure cleanly
- Runs in ~2 GB VRAM at Q4 — fits anywhere
- MIT license, no commercial restrictions
- Optical-compression approach reduces token usage on long documents
Limitations
- 8k context limits multi-page document handling
- OCR-only — not a general-purpose VLM
- Limited reasoning capability beyond extraction
Typical workloads
In our catalog grid, DeepSeek-OCR is filed under Lightweight OCR, Scanned Documents — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: vision-language work — screenshots, charts, scanned documents.
Note the 8k-token context window — fine for short interactions, limiting for long documents or big retrieval contexts. The MIT license is permissive, so shipping it inside a commercial product raises no special legal questions.
Architecture & training
Architecture: Dense vision · 3B · DeepSeek-OCR · specialized in document reading
Training: DeepSeek — massive OCR fine-tuning on scanned documents, receipts, and LaTeX formulas.
Drop-in MIT OCR engine that beats far larger general VLMs at extraction tasks.
Quick start
Install the runtime for your system: Windows, macOS or Linux. Check the exact model tag or GGUF quantization below; catalog IDs are not necessarily Ollama tags.
Start at 4096 tokens of context, then use ollama ps to check GPU/CPU placement. A default download may use a different quantization from the configurator’s memory estimate. Keep the free setup working before considering a kit.
ollama run deepseek-ocr:3bOr use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.
Your private ChatGPT, free, on your own machine in an hour — LM Studio, Ollama, Open WebUI, your documents, no cloud.
- Lifetime online access
- PDF + files
- 30-day refund
Similar models worth comparing
Frequently asked questions
How much VRAM does DeepSeek-OCR need?
At the recommended Q4_K_M quantization, DeepSeek-OCR needs about 2 GB of VRAM. Q8_0 takes 4 GB, and unquantized FP16 weights take 6 GB.
Can DeepSeek-OCR run without a GPU?
Yes — with roughly 5 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.
What context window does DeepSeek-OCR support?
DeepSeek-OCR supports a 8k-token context window (8,192 tokens).
Can I use DeepSeek-OCR commercially?
Yes. DeepSeek-OCR is released under MIT, a permissive open-source license that allows commercial use, modification and redistribution.
How fast is DeepSeek-OCR on consumer hardware?
Our compatibility engine estimates on the order of 70 tokens/sec on a mid-range GPU and up to 160 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.
Which quantization of DeepSeek-OCR should I download first?
Start with Q4_K_M (2 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. It already fits an 8 GB card at FP16.