Skip to main content

​
How Moonshot works

Moonshot AI publishes the Kimi family of models (Kimi K2, K2 Thinking, K2.5). Kimi is particularly strong on long-context code understanding and agentic tasks, and the API is fully OpenAI-compatible, so Kodus talks to it with the existing OpenAI-compatible adapter.
Moonshot’s consumer subscriptions (Andante, Moderato, etc.) apply to the Kimi.ai chat product, not to the API. API usage is billed pay-per-token regardless of any chat subscription you hold. This is different from providers like Z.AI, where the subscription grants API quota directly.
Moonshot operates two parallel platforms:
  • International — platform.moonshot.ai — base URL https://api.moonshot.ai/v1. Billed in USD. This is the default for users outside China.
  • China — platform.moonshot.cn — base URL https://api.moonshot.cn/v1. Billed in CNY. Use only if you already operate inside mainland China.
For the most up-to-date catalog and prices, see platform.moonshot.ai.
ModelPricing (1M input / output tokens)Context WindowNotes
Kimi K2.5 recommended0.60/0.60 / 2.50~256k tokensLatest generation, strongest coding results.
Kimi K2 Thinking~0.60/0.60 / 2.50~256k tokensReasoning variant; use when review needs deeper analysis.
Kimi K2 (0905)lower pricing tier~128k tokensStable general-purpose model.
All three are accessible from the same base URL with different model names:
  • kimi-k2.5
  • kimi-k2-thinking
  • kimi-k2-0905
  • kimi-k2

​
Creating an API Key

A Moonshot account is required to create an API key.
  1. Go to platform.moonshot.ai and create an account (or sign in). Users inside mainland China should use platform.moonshot.cn instead.
  2. Add a payment method — Moonshot does not offer a mandatory free trial on the international platform; a small credit balance may be granted when you first add billing.
  3. Open the API Keys section in the console.
  4. Click Create API Key, give it a descriptive name (e.g. kodus-prod), and copy the key immediately.

​
Configure Moonshot in Kodus

The primary flow is BYOK on Kodus Cloud — you paste the Moonshot key into the web UI and you’re done. Self-hosted users who prefer fixing the provider at the process level can use environment variables instead.
  1. In the Kodus web UI, open Settings → BYOK and click Edit on the Main model (or Fallback, if you want Kimi as a backup).
  2. Toggle the form into Custom mode so you can enter a base URL and a free-text model name.
  3. Fill the fields:
    FieldValue
    ProviderOpenAI Compatible
    API Keyyour Moonshot API key
    Base URLhttps://api.moonshot.ai/v1 (international)
    https://api.moonshot.cn/v1 (mainland China only)
    Modelkimi-k2.5 (recommended) — or kimi-k2-thinking, kimi-k2-0905, kimi-k2
    Max Concurrent Requests5–10 is a good starting point; Moonshot has generous per-account limits
    Max Input Tokensyou can raise this for K2.5’s 256k window; leave default otherwise
  4. Save. Kodus validates the key against the endpoint and surfaces any 401 / 404 immediately.
  5. Open any PR to trigger a review and confirm Moonshot is now serving responses — the BYOK status badge in Settings turns green on the first successful call.
Kimi K2.5’s long context shines on large PRs and on code that spans many files. If your reviews are mostly small PRs, kimi-k2-0905 is faster and cheaper with essentially the same quality for that workload.
You can configure Moonshot as Main and keep an OpenAI or Anthropic key as Fallback — if Moonshot returns 429 or 402, Kodus fails over automatically.

​
Option 2 — Self-hosted (environment variables)

If you run Kodus in Fixed Mode (single global provider, no per-org BYOK), configure Moonshot in the .env of your API + worker containers: 
# Moonshot (Kimi) configuration (Fixed Mode)
API_LLM_PROVIDER_MODEL="kimi-k2.5"
API_OPENAI_FORCE_BASE_URL="https://api.moonshot.ai/v1"    # use .cn for the China platform
API_OPEN_AI_API_KEY="your-moonshot-api-key"
This path is only needed for self-hosted Kodus installs that deliberately disable BYOK. If BYOK is enabled on your self-hosted instance, prefer Option 1 — the UI-based flow is the same as on Cloud.
Restart the API and worker containers after editing .env, then verify the integration: 
docker-compose logs api worker | grep -iE "moonshot|kimi"
For the full self-hosted setup (domains, security keys, database, webhooks, reverse proxy), follow the generic VM deployment guide and only swap the LLM block for the one above.

​
Choosing between Moonshot and an aggregator

You can also reach Kimi K2.5 through OpenRouter, which is already supported as a first-class Kodus provider. The trade-off:
  • Direct Moonshot API — lowest per-token cost, largest context window, native rate limits. Requires a Moonshot account and billing.
  • OpenRouter proxy — single billing relationship for many providers, but a small routing markup and slightly slower cold starts.
Pick direct Moonshot when Kimi is your primary model and you want the best cost and context. Pick OpenRouter when you’re mixing many providers and would rather keep one bill.

​
Troubleshooting

  • Confirm the key is active in the Moonshot console.
  • Verify the base URL matches the platform the key was created on — keys from platform.moonshot.ai do not work against api.moonshot.cn and vice versa.
  • Make sure there are no trailing spaces or quotes in the .env value.
  • Moonshot bills pay-per-token with no mandatory subscription. If balance runs out, requests return HTTP 402.
  • Add funds in the billing section of the console or set a monthly cap to avoid surprises.
  • Confirm the model name matches the catalog (kimi-k2.5, kimi-k2-thinking, kimi-k2-0905, kimi-k2).
  • New model versions are released regularly — check the Kimi docs for the current list.
  • First call after a period of inactivity may cold-start on Moonshot’s side.
  • If latency matters, prefer kimi-k2-0905 over the thinking variant for routine code reviews.
  • Users outside China should always use api.moonshot.ai; api.moonshot.cn may be unreachable or rate-limited from outside mainland China.
  • Confirm outbound HTTPS to api.moonshot.ai is allowed from your Kodus deployment.