DeepSeek-VL
Towards Real-World Vision-Language Understanding
DeepSeek-VL is an open-source vision-language model release for real-world vision and language understanding. The official release includes 1.3B and 7B base/chat variants with general multimodal understanding capabilities across images and text.
Pricing
Official site · Jul 2, 2026, 7:00 AM
Pricing notes were collected, but there are no normalized numeric fields to display yet.
No public pricing is listed in the official DeepSeek-VL repository or associated official materials I could access; treat as unpublished/no public price found.
I could not find any official public pricing page for DeepSeek-VL. The official DeepSeek-VL repository presents it as a model release rather than a paid SKU, and no public price is listed there.
Model Intelligence
Context window
4,096 tokens
Arena ranking
5
Benchmarkable
Yes
Model level
release
Intelligence Index
5.1
MATH-500
0.74
Recent stories
0 linked stories
No linked stories yet.