DeepSeek-V4-Flash-Vision-Exp Release: Multimodal API Now Live
If your business already uses DeepSeek's API for text tasks, this experimental model lets it also read and reason about images — receipts, screenshots, product photos — in the same conversation at standard pricing, which could remove the need for a separate vision-model integration for simple document or image review tasks.
What changed
Source summary
DeepSeek released an experimental multimodal model, DeepSeek-V4-Flash-Vision-Exp, which adds image understanding on top of V4-Flash's existing text, agent, and reasoning capabilities, with multimodal agent performance approaching Claude Opus 4.8. It is available via the Chat Completions, Messages, and Responses API endpoints, accepts mixed text and image input (base64, URLs, or the new Files API), images tokenize at up to 384 tokens each, and it is offered at standard V4-Flash pricing with no extra cost for the Files API.
Why it matters to your work
Practical interpretation
If your business already uses DeepSeek's API for text tasks, this experimental model lets it also read and reason about images — receipts, screenshots, product photos — in the same conversation at standard pricing, which could remove the need for a separate vision-model integration for simple document or image review tasks.
Source & availability
This is a summary of a published provider announcement, with practical context from SCALA’s feed.
Plan, region, workspace and rollout conditions may affect access in Australia. Check the original source for current availability.
api-docs.deepseek.comThis update is primarily for technical teams.