OpenAI built BioBasicsBench, a test that gives AI agents the messy raw data of real lab…
This is a AI post classified by AGK Intelligence as Frontier models (a model release), kept by the AI Radar because it carries real work, not commentary.
OpenAI built BioBasicsBench, a test that gives AI agents the messy raw data of real lab work, like gel photos and microscope videos, and grades whether they read it correctly. GPT-6 Astra earned 66.7% of available credit on a new wet-lab benchmark, versus 43.8% for GPT-5.6 Sol. BioBasicsBench, built with enzyme-design company Ligo, draws 100 tasks from scientists’ raw gels, colony plates, microscopy, sequencing reads and bench videos. Its rubrics give partial credit, including for spotting errors or saying evidence can’t support conclusions. At the xhigh setting, Astra used 4,704 output to
Posted by Rohan Paul (160.8k followers) 25 min ago · 3 likes · 947 views · view the original post on X. Kept by the AI Radar as Frontier models. Tools mentioned: openai.com.
More AI work like this
- Congratulations to Qwen3.8-27B, it beats all models including the top SOTA models from… — @TeksEdge
- GPT-6 Astra caught a wrong antibiotic label on a DNA-cloning control plate that GPT-5.6… — @choblin29
- Inkling 2 Small from @thinkymachines on ARC-AGI (Verified): — @arcprize
- 4B Nemotron 3 Nano beats Frontier models on task specific workflows 🤯 — @chadwahl
- Just ran our evals on Particle-1 — came out 60% cheaper than DeepSeek 4.1 Flash for the… — @Altimor
- In AA-PersonalityBench, Qwen3.8 27B writes the longest responses of any model measured,… — @ArtificialAnlys
- for everyone who where expecting gpt-6.2-astra but disappointed once again today — @TimJayas
- Thinking Machines has released Inkling 2 Small, a model that can compete with much… — @davidarngar
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 65.2k posts from 5k X accounts over the last 14 days, 10.5k tools, 19 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-06 20:15 UTC. Full method.