Thursday, August 13, 2026 · v1.23.0
Model selection and WeCom support
This release lets you choose an AI model and thinking intensity for each conversation. A WeCom app and Custom Alerts expand collaboration and alert handling, while chat attachments, documents, and knowledge outputs are stored together.
New Feature: Model selection
When starting a conversation, you can choose a model and thinking intensity that fit the task. Lightweight work can prioritize fast responses, while complex analysis can receive more thorough reasoning support.
The conversation home now offers DeepSeek V4 Flash, DeepSeek V4 Pro, and Qwen 3.5 Plus, so you can select the right model for each task.
1 New Integration
- New: WeCom app. Create groups, view group and member details, and send direct or group messages to bring communication into task collaboration.
Full Release Checklist
feat
- Added AI model and thinking-intensity selection when starting a conversation.
- Added a WeCom app for creating groups, viewing group and member details, and sending messages.
- Added Custom Alerts for receiving external alerts through webhooks.
- Expanded alert integration for Datadog, Grafana, AWS, Azure, Google Cloud, Coralogix, and Loki.
- Added unified storage for chat attachments, new documents, and knowledge outputs while keeping historical content available.
bugfix
- Fixed an issue where mentioning the bot in a group chat produced duplicate replies after a session was created.
- Fixed an issue where clicking a knowledge reference at the end of a conversation showed the tool execution details panel instead of the knowledge content.