Nace AI launches Drex for Search, cutting agent steps 62% and tokens 56%
testingcatalog · x · 2026-10-26
Nace AI launched Drex for Search, giving agents page-level precision retrieval over cloud files.
- How it works: given a question, it locates the likely folder, then the file, then the page, returning the answer, the quote, and the page number
- Data sources: connects to AWS S3, Azure, and DigitalOcean; reads scanned PDFs, spreadsheets, email, and audio
- Coding agent integration: Claude Code, Codex, and Hermes Agent can use it as a tool, with 62% fewer steps and 56% fewer input tokens per question
- Benchmarks: across 529 queries from six professional work benchmarks including Databricks' OfficeQA, accuracy rose from 63.1% to 67.1%
Related event: Nace AI Launches Drex for Search with API Console Access(2 posts)→
More from coding & agent
- He benchmarks new models by having them build their own copilot — $28 of compute — bucolucas · 2026-10-27
- Should agents be allowed to change plans mid-task? Reddit debates the drift boundary — wistfulcracker84 · 2026-10-27
- HeyGen + Canva MCP turns slide speaker notes into an avatar presentation video — HeyGen · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Cloudflare Computer adds agent interruption recovery and approval pauses in v0.7.0 — irvinebroque · 2026-10-27
- LLM Tails cuts agent input tokens 80% by making agents read before they write — savadks1818 · 2026-10-27