OpenAI adds cloud agents to Codex
Updates focus on persistent workflows and speed
TL;DR:
- Persistent cloud agents for Codex
- Faster inference with Ultrafast mode
- New Decisions API released
Headline
OpenAI adds cloud agents and faster responses to Codex.
Summary
OpenAI released updates focused on agent workflows. These include agents that run continuously in the cloud, tasks that persist, code review plugins, an updated command line tool, and custom tools in ChatGPT. They also launched GPT-6.1 Sol for coding tasks and cross-app work, a Decisions API, and an Ultrafast mode that speeds up responses.
Analysis
The updates connect the model more closely to actual runtimes and interfaces. Cloud setups and plugins make Codex part of the daily workflow rather than just a helper. The Decisions API pushes agents toward ongoing operation and making decisions at the app level. Cheaper access to GPT-6.1 Sol and quicker responses address cost and speed issues. New endpoints and persistent tasks make things easier for developers. These moves help OpenAI compete with other tools by focusing on reliability and how well it fits into existing setups.
Impact Assessment
- Significance: High
- Categories: Product Launch, Developer Tools, Model Release