BAAI

Models & products

Models

AREX

Trends

• Research-agent evaluation shifts to harder domain-specific benchmarks

Items

1. Jul 24

   BAAI's AREX self-improving research agent

   A deep-research agent alternates an inner evidence-gathering loop with an outer constraint-wise self-audit and an autonomous context-compression tool, released as open-weight 4B and 122B-A10B MoE models competitive on BrowseComp and Humanity's Last Exam.

2. Apr 28

   BAAI's AutoResearchBench

   A benchmark for autonomous literature discovery finds the best models reach only 9.39% accuracy tracing a target paper and 9.31% IoU collecting every paper that meets a condition, far below their performance on general agentic browsing.