OpenAI launches hosted API for long-running multi-agent systems
OpenAI released the Agents API in public beta for all developers, packaging the harness and infrastructure behind Codex into a managed service. It supports tool use, hosted or external sandboxes, persistent sessions spanning hours or days, and parallel subagents, while allowing developers to inspect the open-source harness.
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Why it moved the index
The launch lowers the infrastructure barrier to deploying persistent, tool-using systems that can delegate to multiple agents and operate across long sessions. That materially expands practical autonomy and diffusion, including for consequential workflows. Managed sandboxes and an inspectable harness provide controls, but the released capability and broad availability are the larger immediate development.
Assessment history
-
R1
Toward 76 · confidence 98
Adds OpenAI's dated public launch of managed long-running, tool-using, multi-agent infrastructure for all developers.
11 Sept 2026
Share this page
-
DoomBench assesses “OpenAI launches hosted API for long-running multi-agent systems” as evidence moving toward doom, with magnitude 76 and confidence 98 out of 100 in the autonomy and agency category.
-
The DoomBench assessment of “OpenAI launches hosted API for long-running multi-agent systems” is based on reporting from OpenAI and records the editorial rationale, source quality, attribution, and revision history.
-
DoomBench summarizes “OpenAI launches hosted API for long-running multi-agent systems” as follows: OpenAI released the Agents API in public beta for all developers, packaging the harness and infrastructure behind Codex into a managed...
https://www.doombench.com/news/openai-launches-hosted-api-for-long-running-multi-agent-systems-2026-09-10