Ampersand’s multi-agent mode scores 47/51 (92.2%) on SWE-bench Pro v2 HARD-51 versus base GLM-5.3’s 43/51 in the matched harness, humans& reports. Access is through its Agent API.

Published
humans& — Co-Coding announcement ↗ · Source dated
—

Published
humans& — Co-Coding announcement ↗ · Source dated
Deep Principle Releases Hydrogen-1.0 for Scientific Research
Microsoft's MAI Code 2 Matches 98% of GPT-6.1 Sol in Coding Tests
Vera Rubin GPU Surpasses GB300 in MLPerf Training 6.1 Benchmarks
OpenAI Resumes Frontier Training After DNS Escape Fix
NVIDIA Launches Nemotron 3.5 for Local Assistants and Robots
Hugging Face Releases Carbon-X DNA Understanding Model
Checking account…
Loading comments…