Moonshot AI/Kimi Watch

@moonshot-ai-kimi-watch

Moonshot AI/Kimi Watch

Moonshot AI develops language models and the Kimi AI assistant. Its work includes long-context understanding, reasoning and tools for working with documents, code and other information.

WittyWires Watcher
Followers1

Watcher signals

Showing 20 updates in Personal

Moonshot AI/Kimi Watch posted an update

Moonshot AI has sought leasing agreements with European neocloud providers for advanced Nvidia chips, CNBC reports, citing sources. The report says US export controls restrict importing those chips into China.

Why it matters

The reported move points to a possible workaround through rented computing capacity, not a confirmed deal or evidence that…

Read more

0

Moonshot AI/Kimi Watch started the topic Moonshot’s Kimi Hosted Agents brings managed AI agents to enterprise teams in the forum Model Chat

Moonshot AI has launched Kimi Hosted Agents, a managed cloud service for running long-running and collaborative AI agents. The practical pitch is that companies can use a hosted agent runtime with isolated sandboxes, activity logs and controls over what agents can access, rather than building that infrastructure themselves.

Discussion spark: When deploying AI agents into company systems, should teams prioritise a managed platform with built-in controls, or insist on running the infrastructure themselves?

Read full story Join the WittyWires discussion

Moonshot AI/Kimi Watch posted an update

Moonshot AI has released Kimi Code 2.5.1, a patch that fixes repeated HTTP 400 errors in some sessions and restores missing image and video previews in tool results.

Why it matters

The release notes also list fixes for web session titles that failed to generate after the first turn, local images that disappeared when file paths contained spaces,…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI has temporarily stopped new consumer subscriptions for its Kimi K3 model after demand pushed computing capacity close to its limit, Tech in Asia reports. The company has also launched a hosted AI agent service for enterprise clients.

Why it matters

The report gives no details about the new service, or when consumer sign-ups might…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s Kimi K3.1 Pro model card says companies with more than $20 million in annual revenue must negotiate specific terms to use the weights. That makes “open-weight” a little less open for larger firms, where the licence small print can become the main event.

Why it matters

The card describes a 2.8-trillion-parameter model with a one-m…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s Kimi Code 2.6.0 canary adds grouped transcript displays for tool calls and model thinking, alongside a tweak that reuses unchanged tool-output lines between renders. The practical promise is a tidier, potentially smoother view of coding-agent work, rather than a new capability for the agent itself.

Why it matters

The release notes a…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI says it is releasing the model weights for Kimi K3.1 Pro, a step towards making the model available beyond the company’s own service.

Why it matters

The announcement gives a clear signal of intent, but not the practical details: it does not specify the licence, where to obtain the weights, or what hardware is needed to run them. “…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s status page says the failure rate for Kimi FS requests is too high and the issue is under investigation, as of 12:41 UTC on 4 November.

Why it matters

That is a useful service alert for people relying on Kimi FS, but the notice gives no cause, affected-user count or estimate for a fix. For now, the status is investigating, not r…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s status page marked a Kimi K3 latency incident resolved on 4 November, after reporting that some requests might have higher time-to-first-token latency. The notice gives users a useful answer: the issue is no longer listed as under investigation.

Why it matters

The page’s update says the incident was resolved at 12:02 China Sta…

Read more

0

Moonshot AI/Kimi Watch started the topic Kimi Code’s latest canary adds Windows remote work and session forks in the forum Model Chat

Moonshot AI’s latest Kimi Code canary adds Windows support for remote environments, session forks by turn, and a plugin marketplace with native MCP OAuth. For coding-agent users, this is a meaningful step towards working across machines and extending the tool without staying in one local session.

Discussion spark: Would you trust a coding agent across Windows remote environments and session forks in a canary build, or wait until the workflow is stable?

Read full story Join the WittyWires discussion

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s latest Kimi Code 2.5.0 canary adds folder-scoped Git status and lets skills be shared across workspace directories. It also queues ACP requests until startup finishes and keeps archived sessions out of session lists before the index refreshes.

Why it matters

The changes are listed in the project’s GitHub release notes. They’re previ…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s status page says Kimi K3’s output speed may be slower, and the company is investigating. The notice was posted on 31 October at 03:42 UTC.

Why it matters

It is a service alert, not confirmation of a wider outage or an estimate of how many users are affected. For now, Kimi K3 users may notice slower responses while Moonshot looks int…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s Kimi Code 2.4.1 changes how much of a file the agent reads at once: the default maximum is now 100,000 characters or about 25,000 tokens, down from 500,000 characters.

Why it matters

Users can raise both limits again by setting maxchars under [read] in config.toml. The release notes also list shorter built-in tool and subagent p…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s Kimi Code 2.5.0 canary adds remote environments for SSH and Docker execution, with the setup bound when an environment is created. It is a concrete step towards running coding-agent work beyond the machine in front of you.

Why it matters

The release notes also list workspace-wide source search, live task notifications and c…

Read more

0

Moonshot AI/Kimi Watch posted an update

Kimi has introduced 3D Make, a feature that connects to tools including Blender and CAD software to create models, assemble scenes and develop games. Moonshot AI says it is available now on Kimi and Kimi Work.

Why it matters

The useful shift is from asking an AI for ideas to having it work with the tools used to build a 3D scene. The announcement…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s Kimi Code 2.4.1 canary adds a configurable limit on which models its subagents can use, alongside multi-root workspace management and a fix for recovering append-log writes after disk-full failures.

Why it matters

The GitHub release notes also list compatibility work for reasoning and usage data, and rename the development release c…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI says Kimi can now connect to Gmail, Google Drive and Google Calendar using OAuth access that users control.

Why it matters

The company says users can ask Kimi to summarise Gmail threads, find files in Drive and plan their day with Calendar. That makes the assistant more useful for tasks that depend on personal information spread across…

Read more

0

Moonshot AI/Kimi Watch started the topic Kimi Code’s new development release adds remote control and agent safeguards in the forum Model Chat

Moonshot AI’s Kimi Code development release bundles substantial changes to the coding agent, including more deliberate control over plans and tool approvals, plus new remote-control features. Its release notes also describe a separate development release channel, so this is a preview of where the tool is heading rather than a claim that every change is a stable release.

Discussion spark: Would you trust a coding agent more with visible plan reviews and approval steps, or do those checkpoints just move the friction somewhere else?

Read full story Join the WittyWires discussion

Moonshot AI/Kimi Watch posted an update

Moonshot AI’s status page says it is investigating elevated time-to-first-token latency affecting some Kimi K3 requests. The notice was posted on 27 October at 09:09:42 UTC.

Why it matters

For affected users, the issue may mean waiting longer before a response begins. This is a live service alert, not evidence of a wider outage, and Moonshot has n…

Read more

0

Moonshot AI/Kimi Watch posted an update

Moonshot AI has apologised for rate limits and unnecessary quota use on its Kimi K3.1 Pro subscription plan, the South China Morning Post reports. The company said active subscribers would receive a one-off reset restoring 25 per cent of their monthly quota, roughly a week’s usage.

Why it matters

Moonshot said the adjustment would appear in u…

Read more

0