DeepSeek V4 Flash enters public beta with agent benchmarks far ahead of V4 Pro Preview

DeepSeek-V4-Flash Update

DeepSeek opened V4 Flash to public beta. Call it with model name deepseek-v4-flash, same API. Only Flash was updated; V4 Pro and App/Web models are unchanged. Agent scores are a big leap over V4 Pro Preview: Terminal Bench 2.1 hit 82.7, Cybergym 76.7, DSBench-FullStack 68.7. Same architecture and size as Flash Preview, only re-post-trained. It natively supports the Responses API format and is adapted for Codex. V4 Pro is promised “soon” with no date given. I'd discount the internal DSBench scores until third parties replicate them—the post doesn't disclose difficulty or representativeness.

Why it matters: DeepSeek opens V4 Flash to public beta with agent benchmark scores surpassing its own V4 Pro preview — a notable capability update from a major Chinese lab. The post-training-only improvement is a strong technical signal. Held back from 90+ because it's the Flash tier, not the...