AI
DeepSeek releases V4.1-Flash, a 552B-parameter open model activating 8B per token
DeepSeek has officially released DeepSeek-V4.1-Flash, an MIT-licensed multimodal mixture-of-experts model with 552B total parameters, a one-million-token context, and roughly 8B activated per input token. New pricing for the Flash line took effect at 04:00 UTC on September 10, with OpenCode adding support from 04:00 UTC on September 14, expanding the field of frontier-class open-weight models available to developers.