August 1 News.DeepSeek V4-Flash Official Edition API Public Beta, the developer sets the model name to deepseek-v4-flash to call, and the original API call is the same. This time only the API interface for V4-Flash is upgraded, the V4-Pro API, App and web-end models are not adjusted。
The formal version of the model benchmark test exceeded the GLM-5.2 and the structure and parameter size followed V4-Flash-preview, with only re-training。
V4-Flash native supports Responses API and completes fit for Codex. According to the official update log, the relevant code Agent test uses the upcoming DeepSeek Harness very simple mode to run at max slots, Top_p=0.95 and temperature=1.0。
statement:The content of the source of public various media platforms, if the inclusion of the content violates your rights and interests, please contact the mailbox, this site will be the first time to deal with.