OpenAI announces ChatGPT Work, an agentic system for multi-step tasks across apps and files
The company describes a new agent that can persist across hours-long projects, execute actions in external tools, and deliver finished work from high-level goals.
Trust71
HypeSome hype
1 source · single source
- OpenAI describes a new agentic system called ChatGPT Work designed to persist across multi-hour projects and execute actions across users' apps and files.
- The system aims to turn high-level goals into finished work by autonomously taking steps in external tools.
- OpenAI frames the capability as a partner for 'ambitious work' rather than a single-turn assistant.
OpenAI announced a new agentic system called ChatGPT Work, positioning it as a partner for completing multi-hour projects rather than a single-turn assistant. The system is designed to persist across a project's duration, take action across a user's apps and files, and turn high-level goals into finished work by autonomously executing steps in external tools.
The announcement emphasizes the agent's persistence and cross-application execution, framing it as a shift from short, discrete interactions to longer-running, goal-directed workflows.
- Jul 7, 2026 · arXiv cs.CL
Gemma 4 released with multimodal, encoder-free architecture and reasoning traces
Trust79 - Jul 7, 2026 · Simon Willison — everything
Tencent releases Hy3, a 295B-parameter MoE model with 21B active parameters and 256K context
Trust79 - Jul 5, 2026 · Simon Willison’s Weblog
Newer Anthropic models increasingly emit malformed tool calls for third-party edit schemas
Trust79