{
 "license": "CC-BY-4.0",
 "attribution": "Fide AI, Agentic Cyber Explorer",
 "url": "https://agentic-cyber-explorer.pages.dev/events/caisi-deepseek-v4-pro-evaluation-2026/",
 "asOf": "2026-09-26",
 "id": "caisi-deepseek-v4-pro-evaluation-2026",
 "date": "2026-05-01",
 "datePrecision": "day",
 "title": "CAISI evaluation finds DeepSeek V4 Pro trails US frontier models by about eight months",
 "lane": "capability",
 "kind": "eval-report",
 "summary": "NIST's Center for AI Standards and Innovation evaluated the open-weight DeepSeek V4 Pro model and reported that it lags leading US models by roughly eight months in aggregate capability. On a cyber capture-the-flag benchmark it scored well below GPT-5.5 and Claude Opus 4.6, and CAISI notes its non-public benchmarks show weaker agentic performance than DeepSeek's self-reported results.",
 "whyItMatters": "Tracks how quickly open-weight models approach frontier cyber capability, which governs how long closed-model safeguards buy defenders.",
 "actors": [
  "us-caisi",
  "deepseek",
  "nist"
 ],
 "topics": [
  "capability-evaluation",
  "open-weight-diffusion"
 ],
 "atlas": [
  "model"
 ],
 "artifacts": [
  "claude-opus-4",
  "deepseek",
  "gpt-5-family"
 ],
 "sources": [
  {
   "url": "https://www.nist.gov/news-events/news/2026/05/caisi-evaluation-deepseek-v4-pro",
   "publisher": "NIST",
   "title": "CAISI Evaluation of DeepSeek V4 Pro",
   "date": "2026-05-01",
   "type": "primary",
   "accessed": "2026-09-25"
  }
 ],
 "keyFacts": [
  {
   "fact": "On CTF-Archive-Diamond, DeepSeek V4 Pro scored 32% (imputed via item response theory), versus 71% for GPT-5.5 and 46% for Opus 4.6.",
   "locator": "Cyber results"
  },
  {
   "fact": "CAISI estimates V4 Pro lags the frontier by about 8 months, performing similarly to GPT-5; IRT-estimated Elo 800 plus or minus 28 vs 1260 plus or minus 28 for GPT-5.5.",
   "locator": "Key findings"
  }
 ],
 "significance": 3,
 "fideQuestions": [
  "FID-075"
 ],
 "methods": [
  "ctf-benchmarks"
 ],
 "review": "assistant-drafted",
 "addedOn": "2026-09-25"
}