Connect Kimi Code

3 minute read · Chats & Local Agents

Connect Kimi Code, then choose it for a conversation. Your Vortex sign-in and provider sign-in are separate accounts.

• Connect the account
• Check the provider account
• What to expect from Kimi Code
• Confirm a working Kimi conversation
• Account allowance and model choices
• Recover from an interrupted run

Connect the account#

1. Open Settings → Integrations and select Kimi Code.
2. Check installation status. If needed, choose Install and follow preparing, installing and verification progress. An error or cancelled installation needs attention before a run can start.
3. Choose Sign in. Complete the supported browser flow and any displayed provider code, then return to Vortex to verify the account status.
4. Start a new chat, use the provider icon and model picker to choose Kimi Code and an available model. Ask it to explain the first step of your app idea.

Check the provider account#

This integration uses its current provider login. Multiple named account profiles are available for Codex and Claude Code; do not assume the same profile controls are offered for every provider. Check the account connection and model before starting work. See provider accounts, models and usage.

What to expect from Kimi Code#

Connect the appropriate Moonshot AI account and select an available model. An absent quota report appears as Limits not reported; your provider still controls the allowance. Do not treat installation success as proof that the account is connected.

If the app reports unknown or signed out, open the connection controls again. An unknown state is not a confirmed sign-in. A provider account rejection needs an eligible account or another supported agent; Vortex does not change provider eligibility or include its tokens.

Continue with your first app, model and usage choices and permissions.

Confirm a working Kimi conversation#

Complete the installation and provider sign-in stages, then select Kimi Code in a fresh project chat. Start with “Inspect this app and tell me how to run it; do not edit yet.” Check that it identifies the intended project. Ask for a small visible improvement next, then review the run, any approval and the actual Preview result.

An existing executable should be detected and preserved where supported. A sign-in dialog may require a browser or a provider code; complete the displayed flow rather than copying a machine-specific shell command into a conversation. Installation availability and confirmed account access are distinct states.

Account allowance and model choices#

Use the models and supported custom choices exposed by the Kimi integration. Provider-specific controls need not match those of another agent. When Kimi's protocol does not report applicable quota windows, the usage popover says Limits not reported. This is missing information, not a guarantee that the account is unlimited.

Your Moonshot AI account supplies provider access. A provider rejection should be resolved through the provider connection and account. Vortex is free and does not supply a Moonshot AI token budget.

Recover from an interrupted run#

Inspect the last completed step and any modified files before retrying. Stop ends the current operation and invalidates pending approvals. A new run must receive its own approval where required. Follow-up messages can wait for the next turn when active steering is unsupported; check their state before sending them again.

When asking for help, include the visible installation or run stage, model and sanitized error. Do not include provider codes, passwords or full private logs. See permissions and conversation lifecycle for continuing the same work.