# OpenAI introduces BioBasicsBench to test AI agents on raw wet-lab data

> Original title: OpenAI built BioBasicsBench, a test that gives AI agents the messy raw data of real lab work, like gel photos and microscope videos, and ...

- Source: Rohan Paul
- Published: 2026-11-06T19:54:08.000Z
- Bad Influence: https://badinfluence.news/items/px7stnf0p044b2s7w7wdwgbwv
- Original: https://x.com/rohanpaul_ai/status/2118794080885051683

## Summary

OpenAI introduced BioBasicsBench, built with enzyme-design company Ligo, to test whether AI agents correctly read raw lab data such as gel photos and microscope videos. The 100 tasks give partial credit for spotting errors or saying evidence cannot support a conclusion. The post reports GPT-6 Astra scored 66.7% of available credit against 43.8% for GPT-5.6 Sol, and that Astra used about half the older model's output tokens per task at the xhigh setting.

## Post

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 tokens per task, about half the older model’s, so longer reasoning didn’t drive the gain.

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWd3zssaMAAPIiB.png&mode=thumb&exp=1794182400&sig=1c32e1130517f806)

## Quoted post @OpenAI

> A new way to measure how AI agents handle the routine analyses of wet-lab science—BioBasicsBench.
> 
> Developed with Ligo, a biotech company focused on enzyme design, the benchmark draws on 100 tasks from real experiments to test whether agents can interpret raw lab data, identify errors, and recognize when the evidence doesn’t support a conclusion.
> 
> Accelerating biological research starts with getting the everyday work right. BioBasicsBench shows where models need to improve, and helps us track what those gains mean for scientific progress, biosafety, and biosecurity.
> 
> https://openai.com/index/measuring-how-ai-agents-handle-wet-lab-science/

https://x.com/OpenAI/status/2118782708767965238