Prisoner

dev ~3x price increase. Btw, v4-pro-0813 is soooo bad. 🤢 https://x.com/deepseek_ai/status/2087864585504305397

We’re launching DeepSeek-V4-Pro today! 🚀 🔷 Major Agent upgrades with strong production gains! 🔷 Flexible reasoning effort for V4-Pro & V4-Flash: low for simple tasks, high for daily Agent workflows, max for complex tasks. 🔷 Native OpenAI Responses API support, optimized for Codex with one-click setup. V4 Pro is now available on app/web. Try it via “Expert Mode”. V4 Pro is also available via API. Model names remain unchanged—please refer to the API docs for setup details.

We’re launching DeepSeek-V4-Pro today! 🚀 🔷 Major Agent upgrades with strong production gains! 🔷 Flexible reasoning effort for V4-Pro & V4-Flash: low for simple tasks, high for daily Agent workflows, max for complex tasks. 🔷 Native OpenAI Responses API support, optimized for Codex with one-click setup. V4 Pro is now available on app/web. Try it via “Expert Mode”. V4 Pro is also available via API. Model names remain unchanged—please refer to the API docs for setup details.

XDeepSeek (@deepseek_ai)

DeepSeek V4-Pro-0813 w/ DeepSeek Harness (@deepseek-ai/dsh) akari_anya

Legal channels — lowest feasible price

Price (cheap → expensive): Anthropic → OpenAI → DeepSeek → GLM → Kimi

Performance (good → poor): Anthropic → OpenAI → Kimi → DeepSeek → GLM

ちなみに the flash version passed that token test.

DeepSeek V4-Pro-0813 < Claude Sonnet 4.6 < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < ⋯ < Claude Fable 5 akari_anya