Rumored DeepSeek Flash-Cyber benchmarks leak, claiming ExploitBench 96.8
teortaxesTex · x · 2026-10-30
A rumored (self-described as "dreamed," unverified) benchmark table for DeepSeek's Flash-Cyber model is circulating: CyberGym 92.3, ExploitGym 25.4, ExploitBench 96.8, SEC-Bench Pro 71.8 — beating Claude Opus 5 and GPT 5.6-Sol on several cyber evals.
teortaxesTex calls the scores plausible, arguing 4.1F plus targeted RL could trivially reach them, noting DeepSeek already reported CyberGym/ExploitBench/SEC-Bench Pro for V4.1F-0910 and that the 1017 build dropped cyber evals. Footnotes show ExploitBench used a 3000-turn budget vs. competitors' cited 300-turn results. Treat as unverified.
Related event: Rumored DeepSeek Cyber Model Benchmarks Surface Online(4 posts)→
More from Models
- LLMs are fluent in others' work but unreliable when pressed on details — AlexTensor · 2026-10-31
- 576GB 'go f*ck yourself Sam Altman' mirror spreads as OpenAI docs leak — retr0jirachi · 2026-10-31
- Codex 28-day challenge teases day-28 reset: OpenAI may be saving a new model for the finale — haider1 · 2026-10-31
- GPT-6 ASTRA LAW spotted in OpenAI's API model list, launch looks imminent — imjustnewatai · 2026-10-31
- Grok now accepts Alipay and UnionPay, opening paid access for Chinese users — op7418 · 2026-10-31
- Asking DeepSeek and Hy to predict their own next models yields two very different roadmaps — lucky-adventurer- · 2026-10-31