Evaluating AI’s ability to perform scientific research tasks

What changed

OpenAI introduces FrontierScience, a benchmark testing AI reasoning in physics, chemistry, and biology to measure progress toward real scientific research.

Why it matters

A concrete addition to Practical AI: Tools, Models & Frameworks: it changes what's available to builders today rather than being general commentary.

How it compares

Related prior coverage to compare against:

• Introducing GeneBench-Pro
• PaperBench: Evaluating AI’s Ability to Replicate AI Research
• Early experiments in accelerating science with GPT-5

• Introducing GeneBench-Pro similarity 0.32
• PaperBench: Evaluating AI’s Ability to Replicate AI Research similarity 0.30
• Early experiments in accelerating science with GPT-5 similarity 0.28