# Kimi K3.1 Pro and team usage

Source: https://www.hyperagent.com/docs/changelog/2026-11-13-model-and-team-usage

> For AI agents: the documentation index is at https://www.hyperagent.com/llms.txt and the complete documentation in one file at https://www.hyperagent.com/llms-full.txt. Any docs page is also served as Markdown by appending `.md` to its URL.

Kimi K3.1 Pro, team usage by model and person, and source-linked room recaps.

## Kimi K3.1 Pro [#kimi-k31-pro]

Select Kimi K3.1 Pro in Manual when it appears in your model picker. Its availability depends on your account and Team policy. See [Models and compute](https://www.hyperagent.com/docs/concepts/agents/models).

## Team usage by model and person [#team-usage-by-model-and-person]

Compare your Team's recorded usage by model or by person. Owners can inspect each person's share. Other members see their own usage and a combined total for everyone else. Some recent or older usage may remain unattributed, so an unavailable total is not zero. Open [Usage and invoices](https://www.hyperagent.com/docs/billing/invoices).

## Evaluate shared work [#evaluate-shared-work]

Evaluate an existing response in a Team-owned thread with a rubric you can access. The evaluation scores that response. It does not run the task again. Owners and Editors can pin or unpin rubrics for the shared thread. See [Rubrics and evaluations](https://www.hyperagent.com/docs/learning/rubrics-and-evaluations).

## Read a room's daily recap [#read-a-rooms-daily-recap]

Open a room's Recap view to generate or read a recap for a UTC date. Follow a citation to open its source message. A new or revised recap counts toward usage. Saved recaps do not consume usage when you open them.

Other confirmed updates include:

* **Media outputs.** Choose Fast, the default, or High quality for narration. Cast supported stored videos and download media with its original extension when available.
* **Databricks defaults.** Queries use the configured catalog and schema. Discovery follows **Default Schema**. Choose a catalog if the defaults cannot resolve one.
* **Live check models.** Defaults follow the agent's model family when the thread can run it. Eligible Auto checks may use a connected ChatGPT subscription for supported models.
* **Work history.** Automation History shows where incoming events stopped. **See complete work record** opens the conversation’s ordered actions and available step **Input** and **Result**.
* **Squad shortcuts.** Open a squad from the sidebar. Where it has a Home room, choose **Message** to start or continue a conversation.
* **Rubric lifecycle.** The editor blocks invalid weights before saving. Archive eligible personal rubrics before permanent deletion. Restore and attach them again when needed.