News brief · About 2 min read
Kimi Code Desktop 1.1.1 adds workspace trust and clearer session controls
The October 21 desktop update adds trust checks, model-switch choices, queued compaction and session reload, alongside workflow fixes.
- Published here
- Source published
- Source checked
Kimi Code’s official changelog dates Desktop 1.1.1 to October 21, 2026. The short entry describes controls for workspace trust and ongoing sessions, plus interface and reliability fixes. This records an earlier desktop update, not a new Kimi foundation model or a release first announced today.
Listen to the news
Read this article aloud or keep listening to the news.
Preparing the reader…
Device voice
Speech is handled by your browser or device. AISubHub does not send the text to an audio API. Voice availability and playback with a locked screen vary by device.
Resuming may repeat a short phrase.
At a glance
- Opening an untrusted folder now asks whether to trust it; declining blocks sending messages. Switching models mid-session offers three choices: start a new session, compact and continue, or continue directly.
- While the Agent is running, /compact can be queued and executes when its turn comes; it does not mean immediate compaction or interruption. /reload or the session menu immediately applies external changes to AGENTS.md, skills and configuration.
- A session already occupied by another client shows a confirmation with one-click retry. The File Tree can sit on either the left or right, using a new toggle icon.
- Kimi reports fixes for stutter and device heating while streaming long conversations; loss of message text, attachments and queued messages after send failures; and the Windows kimi app failing to open workspaces or wake the desktop app, as well as new sessions stuck loading.
Why it matters
AISubHub’s reading is that these changes make choices about trust, model changes and applying external edits more explicit within a working session. Queued compaction and immediate reload are distinct behaviors; the changelog does not establish a measured productivity gain.
Limits and availability
We read the complete 1.1.1 entry, not the entire release history, and did not install or test the app. The safeguard is not an independent security or sandbox audit; occupancy confirmation is not a distributed-locking guarantee. Reported fixes do not guarantee no further data loss, heating or loading failures. Prices, plan entitlements, regional availability, minimum OS versions and any CLI update are not established by this entry.
Sources and evidence
Source published: · Source checked:
AI-assisted editorial articles checked against linked primary sources. Vendor claims are attributed; inclusion is not endorsement.