We’re launching DeepSeek-V4-Professional at present! 🚀
🔷 Main Agent upgrades with robust manufacturing positive factors!
🔷 Versatile reasoning effort for V4-Professional & V4-Flash: low for easy duties, excessive for every day Agent workflows, max for complicated duties.
🔷 Native OpenAI Responses API help, optimized for Codex with one-click setup.
V4 Professional is now accessible on app/internet. Attempt it through “Skilled Mode”.
V4 Professional can be accessible through API. Mannequin names stay unchanged—please discuss with the API docs for setup particulars.
API pricing replace 💰
With the V4 lineup launch, we’re updating our API pricing and introducing peak and off-peak charges. Off-peak charges are 50% decrease than peak, enabling extra versatile workload scheduling. 📉
New pricing takes impact at 16:00 UTC, Aug 16, 2026 🕒
