DeepSeek V4 Pro
An AI model from DeepSeek available through an API, using a MoE architecture and supporting long-context coding and agent tasks.
What we have written
- DeepSeek V4 Pro turns model routing into the real eval2026-09-01
DeepSeek V4 Pro sits in a new evaluation pattern: the model card still matters, but the harder question is whether the right model can finish a tool-using job at an acceptable cost and speed.
- DeepSeek V4 Flash turns AI work into a routing problem2026-08-15
DeepSeek V4 Flash makes the default AI step look less like a premium model call and more like a routed, high-volume operation. geekpark says the official version launched on July 31, while woshipm points to agent-facing scores of 82.7 on TerminalBench 2.1, 76.7 on Cybergym, and 70.3 on Toolathlon Ve
Where it came from
The Chinese reporting these articles are built from. Read in Chinese; the facts are ours to restate, the words are not.
- 最前线｜本地大模型 StartLux-27B 通过 MCP-Universe 测评：综合第二，多项专项第一36kr
- 实测正式版 DeepSeek V4 Pro，补齐 Agent 能力｜AI 上新geekpark
- DeepSeek 上线多模态，我用它做了《牛来》小游戏｜AI 上新geekpark
- 实测 DeepSeek V4 Pro 正式版：能力直逼 Fable 5，还藏了一个大招ifanr
- DeepSeek V4 Pro 正式版上线，多项指标逼近 Fable 5ifanr
- 3美分打3美元，DeepSeek V4 Flash 杀穿的不是模型，是定价woshipm
- DeepSeek V4-Flash火了：AI产品经理真正该学的，不是追模型，而是做“模型分层”woshipm
- 3分钱一个任务，B端该怎么选woshipm
- 一夜之间，DeepSeek V4 Pro和Grok 4.6把全球大模型推入了新的斩杀线。woshipm
- DeepSeek V4 Pro 遇上 Grok 4.6，新一轮模型竞争开始 比什么？woshipm