DeepSeek-V4.1-Flash launches with native visual understanding
DeepSeek has released V4.1-Flash on its API with built-in visual understanding. It replaces the previous Flash models, and DeepSeek says API prices have been reduced.
What changed
V4.1-Flash is the smallest model in DeepSeek's new architecture family and includes native multimodal visual understanding. Developers can access it using the model name `deepseek-flash`. V4 Flash and V4 Flash Vision Exp have been retired, with their API names temporarily routed to V4.1-Flash. DeepSeek also announced reduced API prices, linking to its pricing page for details.
Why it may matter for your work
Built-in visual understanding could help businesses build tools that work with images as well as text. Lower API prices may reduce running costs, while teams using the retired models may need to check how the replacement behaves in their existing workflows.
What it can actually do
Understand visual inputs through the DeepSeek API using native multimodal support.
Limits to check
The old Flash model names are only temporarily routed to V4.1-Flash; no end date is stated. Exact API prices are not included in this change-log entry.
“DeepSeek V4.1 Flash is now available on the DeepSeek API with native multimodal support.”From the linked source page