DeepSeek-OCR vs Qwen3.5-2B

Specs, VRAM requirements and download trends — updated 10 November 2026.

deepseek-ai · A · Open weights

Params

3.3 B

Context

8,192

30d

2.0 M

mit · commercial OK

Qwen · B · Open weights

Params

2.3 B

Context

—

30d

3.7 M

apache-2.0 · commercial OK

Specification comparison

Differences are highlighted; identical values are muted.

Specification	DeepSeek-OCR	Qwen3.5-2B
Parameters	3.3 B	2.3 B
Architecture	DeepseekOCRForCausalLM	Qwen3_5ForConditionalGeneration
Context length	8,192	—
Licence	mit	apache-2.0
Commercial use	Allowed	Allowed
Languages	1	—
Downloads 30d	1,952,733	3,743,248
Downloads all time	38.7 M	22.3 M
Quantizations on the Hub	SAFETENSORS	SAFETENSORS
Gated	No	No
First seen on the Hub	2025-10-17	2026-02-28

Daily snapshots, last 90 days (13 Aug – 10 Nov)

4.9 M2.0 M

DeepSeek-OCR Qwen3.5-2B

On RTX 4090 · 24 GB · 8K context unless noted

Qwen3.5-2B · fp16 5.8 GB / 24 GB

✅ Runs comfortably

DeepSeek-OCR · fp16 8.3 GB / 24 GB

✅ Runs comfortably