TestingOps.ai

AI model providers · Model provider

Kimi and TestingOps: test agents on your Moonshot AI key

Save a Moonshot AI key and the agents on your TestingOps projects run Kimi models through the OpenCode adapter, on your own Kimi platform account.

  • Model provider
  • Available with limits
Request a demo Kimi API quickstart
Connection
Model provider
Writes back
Nothing, only model requests
Last checked
7 Nov 2026

In short

Kimi and TestingOps in short

The Kimi connection lets TestingOps agents run Kimi models with your own Moonshot AI API key, which the customer admin saves under Integrations → AI models with the preset address https://api.moonshot.ai/v1.

While Kimi is your first provider, your agents move to the OpenCode adapter and call Kimi with your key, on kimi-k2.6 unless you enter another model.

Kimi shares one slot with the other OpenAI-compatible providers, and TestingOps' own AI features outside agent runs do not use a Kimi key.

What it does

What it does for your testing

  1. 1

    Agents on Kimi, billed to your Kimi account

    Input
    Your Moonshot AI key, first under Your providers
    TestingOps
    Moves your organisation's agents to the OpenCode adapter with a moonshotai/<model> model and attaches the key to your projects as MOONSHOT_API_KEY
    Output
    Agent runs on your projects that use Kimi with your key
  2. 2

    Model names the platform offers

    Input
    Your saved key
    TestingOps
    Fetches the model list from api.moonshot.ai for the Model suggestions; Test connection warns when a name is not listed
    Output
    A model name that agents can actually call
  3. 3

    A copy for RunningOps

    Input
    An organisation linked to RunningOps
    TestingOps
    Copies the key and the provider order there as TestingOps · Kimi (Moonshot)
    Output
    The same Kimi account on the execution side

Step by step

How it works

  1. Add KimiThe customer admin picks Kimi (Moonshot) under Add a model provider; Base URL (preset) shows https://api.moonshot.ai/v1. Paste the key and enter a model or keep the default.
  2. Stored and attachedThe key is stored encrypted and attached to every project of your organisation with the address and model.
  3. Kimi first, agents followWhen Kimi (Moonshot) is marked Default · drives agents, your agents switch to OpenCode at their next run and use Kimi through it.
  4. One OpenAI-compatible slotIf you also keep Qwen, DeepSeek, Azure OpenAI, Ollama or a self-hosted endpoint, only the highest-ranked of them is stamped on your projects; the others show Standby.

Moonshot's documentation now lives on platform.kimi.ai; the API address in its quickstart, and in TestingOps, is https://api.moonshot.ai/v1.

Connect

Setup

  1. Create an API key on the Kimi platform's API keys page.
  2. In TestingOps, as customer admin, open Integrations → AI models and pick Kimi (Moonshot) under Add a model provider.
  3. Keep Base URL (preset), paste the key into API key, enter a Model or leave it empty, and select Save.
  4. Select Test connection on the entry, then move Kimi (Moonshot) to the top under Your providers and select Save order.

Documentation: Kimi API quickstart

Where you connect it in TestingOps: Integrations → AI models → Your providers → Kimi

Use cases

Use cases

Run agents on a Kimi contract

Trigger
Your company has an agreement with Moonshot AI
Uses
Your Kimi key, first under Your providers
Agent activity
Agents on your projects work through OpenCode on Kimi
Result
Agent usage billed to your Kimi account

Compare Kimi with your current provider

Trigger
You want to see Kimi's output on your own projects before deciding
Uses
Your provider order
Agent activity
Moving Kimi to the top switches the agents at their next run; moving the previous provider back reverses it
Result
Real work from two providers on the same projects

Move to a newer Kimi model

Trigger
Moonshot releases a model you want to use
Uses
The Model field, filled from the live model list
Agent activity
Agents move to the new model at their next run
Result
A model change without a new key
Condition
Test connection warns if the endpoint does not list the name

Capabilities

Capabilities and limits

What is read

  • Kimi's responses while agents work on your projects
  • The model list at api.moonshot.ai for your key

What is produced

  • Agent runs on your projects that use Kimi through OpenCode
  • A Reachable or Unreachable mark after Test connection

Where it can write

Nothing in your Kimi account; only model requests are sent with your key.

How it connects

Moonshot AI's OpenAI-compatible API at https://api.moonshot.ai/v1 with your key; agents through the OpenCode adapter. A linked RunningOps organisation gets a copy of the key.

Account and permissions

  • A Kimi API key from the Kimi platform console
  • Customer admin rights in TestingOps, or a company admin of the TestingOps workspace managing keys for your organisation

Limits

  • TestingOps' own AI features outside agent runs do not use a Kimi key
  • Shares one OpenAI-compatible slot: only the highest-ranked of Kimi, Qwen, DeepSeek, Azure OpenAI, Ollama and self-hosted is active
  • Agent runs reach Kimi through OpenCode's Moonshot provider; an address other than the preset is not confirmed for agent runs
  • One key per organisation for all of its projects

Example

Example: how a team uses it

Illustrative scenario, not a customer story.

A team evaluating models for test design adds a Kimi key next to its existing Anthropic key and moves Kimi (Moonshot) to the top for two weeks. The agents write that period's cases on Kimi through OpenCode.

At the end of the trial the customer admin moves Anthropic (Claude) back to the top and selects Save order; from their next run the agents are on Claude again, and the Kimi key stays saved for later.

Frequently asked questions

Do we need an OpenAI account for Kimi?

No. Kimi's API accepts OpenAI-style requests, but the key is your Moonshot AI key and the requests go to api.moonshot.ai.

Why does Kimi show Standby?

Another OpenAI-compatible provider is ranked higher. Only one of them is stamped on your projects; move Kimi above it and select Save order.

Do TestingOps' own AI features use the Kimi key?

No. The key is used for agent runs, the connection test and the copy on RunningOps; features outside agent runs use other models.

Which model runs if Model is empty?

kimi-k2.6, TestingOps' preset. Enter another name from the list to move your agents to it.

Next step

Request a demo to try Kimi on a test project with your own Moonshot AI key.

Request a demoMore ai models integrations