Third-party benchmark claims GPT-6 Astra leads frontier models at antibody developability prediction

gdb · x · 2026-09-11

Account @andrewaiginin reports that its benchmark shows "GPT-6 Astra" outperforming other frontier models at predicting antibody developability — whether an antibody will aggregate, stay stable, and behave like a real drug. An interactive antibody visualization was reportedly also built with Astra in about 1 hour. Note: neither the model name nor this third-party eval is officially confirmed.

Original post →

More from Models

Models channel →