OpenAI Agents API ships 12-hour prompt cache, async functions, and per-session budget control
OpenAIDevs · x · 2026-11-10
OpenAI's Agents API team shipped three updates:
- 12-hour prompt cache: prompts stay cached for 12 hours, cutting costs for long-running agents, especially personal agents.
- Async functions: agents keep working after a tool call without waiting for results — useful for starting slow lookups early and answering independent parts of a request.
- Budget control (beta): per-session limits and monitoring of agent token spend, so one customer can't burn through all your tokens; access currently via the team.
The team is also soliciting feature requests for the coming weeks.
More from coding & agent
- Painting with Claude Fable 5.2: each artwork is a 1000-line Python brush engine — joellewenstein · 2026-11-10
- MCP adds Events: agents subscribe to deployment.succeeded instead of polling — cramforce · 2026-11-10
- Two-model workflow: Fable 5.2 for planning, Opus 5.5 for execution — daniel_mac8 · 2026-11-10
- Replit launches Monitoring Agent to auto-detect, fix and deploy fixes for vibe-coded apps — amasad · 2026-11-10
- Multi-agent harnesses keep beating single agents on accuracy and latency in ProgramBench — marktenenholtz · 2026-11-10
- fx: a tiny open-source coding agent in Zig with a 5.97 MiB single binary — evilrabbit_ · 2026-11-10