BioBasicsBench: 100-task benchmark for AI wet-lab skills
In brief
OpenAI released BioBasicsBench, a 100-task benchmark to evaluate how well AI agents interpret raw experimental data and perform everyday wet-lab science tasks.
How it works
The benchmark consists of 100 tasks that test an AI agent's ability to understand raw experimental data and execute common wet-lab procedures. Agents are evaluated on their accuracy in interpreting results and handling practical lab scenarios.
Researchers and developers building AI agents for biological laboratory work.
Availability
Sources
1 merged into this cardMore from OpenAI
Codex TUI WebRTC Realtime Voice Calls
Codex v0.170.0 adds WebRTC negotiation, allowing users to conduct realtime voice calls directly in the terminal user interface.
Codex: Refreshable OpenTelemetry HTTP Headers
Codex now automatically refreshes authentication headers for OpenTelemetry HTTP exporters, ensuring continuous telemetry without token expiration issues.
Codex v0.170.0: Multi-Environment MCP Support
Codex version 0.170.0 adds experimental support for using identically named Model Context Protocol servers across different environments.