Kimi K3 in Kilo Code

Kilo CLI: built in since 7.7.5 · VS Code and JetBrains: OpenAI-compatible provider · about 2 minutes · CLI listing verified on 7.7.5 (19 Sep 2026)

Before you start. Create an account (you get $2 to try), then open Token Management → Create token and copy the key. The key starts with sk-. Endpoints: OpenAI-compatible https://microquickjs.com/v1, Anthropic-compatible https://microquickjs.com. Model ID: Kimi-K3.

Kilo CLI: pick it from the list

Kilo CLI ships the models.dev catalog, and ainetcafe has been in it since 17 September. From Kilo CLI 7.7.5 (published 18 September) there is nothing to configure: export the key and the provider appears.

export AINETCAFE_API_KEY="sk-..."   # add to your shell profile
kilo models ainetcafe               # prints: ainetcafe/Kimi-K3
kilo --model ainetcafe/Kimi-K3

Inside the TUI, /models lists ainetcafe / Kimi K3 once the variable is set; without it the provider is hidden, which is how Kilo treats every key-based provider. kilo models ainetcafe --verbose shows what the catalog carries for us: $2.10 / $10.50 per million, $0.30 cached input, 262,144 context, 32,768 output, reasoning and image input on. To pin it in a project, add to kilo.json:

{ "model": "ainetcafe/Kimi-K3" }

VS Code / JetBrains extension: OpenAI-compatible profile

Open Kilo Code settings → Providers → add a profile.
Provider: OpenAI Compatible.
Base URL: https://microquickjs.com/v1; API key: your ainetcafe key.
Model: type Kimi-K3 (Kilo may not list it until it has fetched /v1/models; typing the ID is fine).
Context window 262144, max tokens 32768, enable image support and tools.
Assign the profile to the modes you use (Code, Architect, Ask, Debug). K3 works for all of them.

Notes

Orchestrator mode spawns subtasks; each carries its own context, so prompt caching matters more here. Leave it on.
If Kilo shows "reasoning" content, that is K3's thinking block. Collapse it in the UI; do not disable it in the request.

Verify the key first

Thirty seconds now saves an hour of guessing later. Run this in a terminal; you should see a short reply and a usage block.

curl https://microquickjs.com/v1/chat/completions \
  -H "Authorization: Bearer $AINETCAFE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"Kimi-K3","messages":[{"role":"user","content":"Say OK"}],"max_tokens":5}'

If something goes wrong

Symptom	Cause	Fix
401 Invalid token	Key copied with a space, or key from another site	Re-copy from Token Management on microquickjs.com
403 quota / insufficient balance	Wallet at zero or plan quota used up	Top up in Wallet, or wait for the weekly plan reset
404 model not found	Model ID typed differently	Use exactly `Kimi-K3`
429	Rate limit for your tier	Reduce concurrency; limits rise with spend, or ask for a committed plan
Tool calls never happen	Tool calling disabled in the client, or thinking blocks stripped	Keep the client's default tool settings; K3 returns a `thinking` block before `tool_use`, do not filter it out

Still stuck? Email [email protected] with the request ID from the error body.