BioBasicsBench

Reporting on BioBasicsBench from Lazy Tensor.

Search this sectionA place, person or subject

Adaptive AI & Lab Benchmarks

Google Debuts Gemini Decisions API as OpenAI and METR Test Frontier AI

Google AI introduces an experimentation and contextual bandit service for real-time optimization while OpenAI open-sources a 242-question wet-lab biology benchmark and METR evaluates AI R&D speed at Anthropic.

Nov 12, 2026By Lazy Tensor4 min read

An Economist-style conceptual print illustration on warm cream paper showing branching geometric decision paths, a laboratory flask, and a balanced mechanical scale.

Google AI Gemini Decisions API BioBasicsBench