OpenAI introduces BioBasicsBench, a 100-task benchmark for AI agents in wet-lab science
BioBasicsBench is a 100-task benchmark that tests how well AI agents interpret raw experimental data and handle routine wet-lab science.

Watch the short
Under a minute, read by an AI anchor. Made .
YouTube loads only when you press play.
OpenAI has introduced BioBasicsBench, a set of 100 tasks that tests how well AI agents handle everyday wet-lab science. The benchmark asks whether an agent can interpret raw experimental data and carry out routine laboratory work.
The announcement does not name the models tested or report what the agents scored. Without those details, it defines what BioBasicsBench measures but offers no basis for judging current agent performance.
Its two areas are interpreting raw experimental data and carrying out everyday wet-lab work.
What changed
OpenAI introduced BioBasicsBench, a 100-task benchmark testing how well AI agents interpret raw experimental data and handle routine wet-lab science.
Type: Research
Why it matters
Companies that use AI agents for lab analysis now have a dedicated test of raw-data interpretation. That skill differs from answering general science questions.
What you can do
Ask vendors of lab-facing AI agents how their products perform on this benchmark before trusting them with raw experimental data.
Original source
We have no link to the original source for this story yet.
Where we saw it
OpenAI
AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it. How it works