OpenClaw 2.0 has been launched with easier setup, subagent tracking & private credentials

The popular autonomous AI agent OpenClaw has released version 2.0 after nearly two months of development. The update focuses on simplifying installation and setup, with better detection of existing ChatGPT or Claude subscriptions, API keys, and local models, while reducing the number of configuration options shown during initial setup. The browser app has also been rebuilt, allowing users to continue configuring OpenClaw, return to ongoing tasks, and follow agent activity as it happens.

OpenClaw 2.0 adds search for past conversations using exact words or phrases, persistent progress cards that survive reloads, visibility into subagent activity and accumulated edits, and support for carrying out work on paired devices or in the cloud. Agents can now request credentials through masked prompts without exposing them in chat or model context, with an optional proxy limiting protected secrets to approved destinations. Slack support has also expanded with Enterprise Grid identity, direct messages, workspace routing, actions, approvals, and presence, while audio and video handling now includes persistent media, native playback controls, and video uploads on Apple and Android clients.

The release took longer than usual as the team grew and reworked its release process to handle the increased development volume, with additional time spent testing both clean installations and upgrades. As a reminder, OpenClaw creator Peter Steinberger joined OpenAI in February 2026, while development of the project has continued through its independent team.