DeepSeek's new model needs a quarter of the memory to hold a chat.
DeepSeek released V4.1 Flash on Thursday, its first model that reads images as well as text. The 552-billion-parameter model keeps a conversation in a quarter of the fast memory the old one needed. (Correction: the old flagship V4 Pro was due to be retired on 14 September, but DeepSeek is keeping it running.)
Why it mattersLess memory per conversation means cheaper long chats, and prices usually follow. A Chinese lab reaching this level again drags the whole market's prices down.
▶ Related video: DeepSeek V4.1 Flash Is INSANELY GOOD! Fast, Cheap, Powerful! (Fully Tested)
Read in the app — free, in 9 languages
Related stories
Block's new voice AI learned to talk without reading any text.2026-10-25Moonshot's new Kimi trails only three models on an independent index.
2026-10-24A free Chinese image model passed a million downloads in a week.
2026-10-23Canva trained its first image model on templates, not photos.
2026-10-23Hugging Face built a chatbot that reads DNA and explains mutations.
2026-10-22