See Where Your Organization's Credits Go
The new Insights dashboard in organization settings breaks down credit usage across your org. Get an overview, a per-model view of credits and cost, a Credit Explorer for digging into individual runs, and forecasts that project where your spend is heading.
Set Organization Model Presets
Organizations can now choose the models behind the Recommended, Smartest, and Fastest presets, so every agent in the org runs on the models you've picked for each tier.
GLM-5.2 (Z.ai)
Added GLM-5.2 for agents, running under Zero Data Retention.
Agent Improvements
- Agent chat search now looks inside conversations, not just chat names, so you can find a chat by something that was said in it.
- Faster code sandbox startup for agents with large workspaces.
Platform Improvements
- Improved browser tab titles so your tabs are cleaner and easier to tell apart.
- Added filters to artifacts and a getting-started card for live dashboards.
- Running an agent inside a looping workflow now shows the right credit cost for each item in the list.
- Chat credit costs now include background actions like self-improvement enrichment and chat evaluations.
Bug Fixes
- Fixed human-in-the-loop approvals not reaching custom Slack apps. Add an interactivity URL to your Slack app to turn it on.
- Fixed subagent approval requests not surfacing to the parent agent.
- Fixed the "Ask for writes/deletes" approval option showing for read-only apps that can't make changes.
- Restored the option to download a skill from the agent preview.
- Fixed agents continuing to reply in Slack after you delete the message that started them.
- Improved the reliability of Gemini models across long-running conversations.