deepseek-v4-flash

1,809 5 hours ago

DeepSeek-V4-Flash is a preview of the DeepSeek-V4 series, a Mixture-of-Experts model with 284B total parameters and 13B activated, built for efficient reasoning across a 1M-token context window.

tools thinking cloud

1 model

deepseek-v4-flash:cloud
ea027821675c • 1M context window •
Text input • 5 hours ago