// DeepSeek · Vision-language
DeepSeek-OCR-2: VRAM & local deployment
3.39B (MoE) parameters. About 3.3 GB at 4-bit and 8.5 GB at BF16 with one 8K-token request.
released 2026-01-27licence: Commercial OKmemory figures are estimates
01Specs
| Publisher | DeepSeek · deepseek-ai |
|---|---|
| Family / series | DeepSeek · DeepSeek-OCR-2 |
| Type | Vision-language (image-text-to-text) |
| Parameters | 3.39B (MoE) Source: counted from official safetensors weights |
| Context length | 8K tokens (8,192)max_position_embeddings in config.json |
| Architecture | 12 layers · GQA · MoE |
| Released weights | BF16 · 6.8 GBdownload |
| Licence | Apache-2.0 · Commercial OKPermissive licence (Apache-2.0 / MIT / CC-BY / OpenMDW): commercial use allowed. |
| Released | 2026-01-27 (Hugging Face repo created) |
02How much VRAM it needs (estimate)
Total = weights + 10% + 0.5GB overhead + KV cache for one request (FP16 KV). More concurrent users multiply the KV column.
| Precision | Weights | Total @ 8K | Smallest GPU | Mac |
|---|---|---|---|---|
| BF16 / FP16 (original) | 6.8 GB | 8.5 GB | 12GB GPU | Mac 16GB |
| FP8 / INT8 / Q8_0 | 3.6 GB | 5.0 GB | 8GB GPU | Mac 16GB |
| 4-bit (Q4_K_M / AWQ / GPTQ-Int4) | 2.0 GB | 3.3 GB | 8GB GPU | Mac 16GB |
KV cache per request: 8K 0.5 GB
Bits per weight: BF16 16, FP8 8.5, 4-bit 4.85. “Fits” = total ≤ 90% of memory; Mac GPU can use ~2/3 of unified memory (≤36GB) or ~3/4 (larger). For MoE, all experts must be in memory, so size by total parameters. Method →
03How to run it
Only runtimes the publisher or the runtime’s official library names for this model.
- vLLM — mentioned in the official model card
- Transformers — mentioned in the official model card
04Official repositories & quantised builds
1 official repo(s) from deepseek-ai.
- DeepSeek-OCR-2 main
05Other DeepSeek models
All 34 →| Model | Params | 4-bit total | Released |
|---|---|---|---|
| DeepSeek-V4.1-Flash-1017 | 748B / 16B | 499 GB | 2026-10-17 |
| DeepSeek-V4.1-Flash | 748B / 16B | 499 GB | 2026-09-10 |
| DeepSeek-V4-Pro-0813 | 1.65T | 1.10 TB | 2026-08-13 |
| DeepSeek-V4-Flash-0731 | 304.2B | 203 GB | 2026-07-31 |
| DeepSeek-V4-Flash | 284B / 13B | 190 GB | 2026-04-22 |
| DeepSeek-V4-Pro | 1.6T / 49B | 1.07 TB | 2026-04-22 |
| DeepSeek-V3.2 | 685.4B | 458 GB | 2025-12-01 |
| DeepSeek-V3.2-Speciale | 685.4B | 458 GB | 2025-11-28 |
06Sources
Data checked 2026-02-03. Parameters, context and licence are copied from the official repository; memory figures are our estimates.