NewHow we catch a seller listing the wrong model, and refund you if one does.

Kilo Code

Kilo Code adds MartLLM as a Custom provider with the OpenAI Compatible API type.

Fastest: one command

npx uberllm init --tool kilo-code

This connects your account in the browser, saves the key to a file only you can read and adds the MartLLM provider to ~/.config/kilo/kilo.jsonc, with each model's real context and output limits. The file is backed up first, and npx uberllm init --undo puts it back. Details: uberllm init. The steps to do it by hand are below.

Before you start

Setup in VS Code

  1. Open Kilo's Settings (gear icon) and go to the Providers tab.
  2. Scroll to the bottom and click Custom provider.
  3. Fill in:
    • Provider ID: uberllm
    • Display name: MartLLM
    • Provider API: OpenAI Compatible
    • Base URL: https://api.martllm.com/v1
    • API key: your ull_... key
    • Models: click Fetch models and pick the ones you want, or Add model and type an id such as z-ai/glm-5.2.
  4. Click Submit.

The fetched list carries only model ids and names. Set each model's context and output limits in the config file (next section): without them Kilo uses 0 and its context management does not work.

Setup in kilo.jsonc

Global file: ~/.config/kilo/kilo.jsonc. Project file: kilo.jsonc in the project.

{
  "provider": {
    "uberllm": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "MartLLM",
      "options": {
        "baseURL": "https://api.martllm.com/v1",
        "apiKey": "{env:UBERLLM_API_KEY}"
      },
      "models": {
        "deepseek/deepseek-v4.1-flash": { "name": "DeepSeek V4.1 Flash", "limit": { "context": 1048576, "output": 384000 } },
        "z-ai/glm-5.2": { "name": "GLM 5.2", "limit": { "context": 1048576, "output": 32768 } },
        "moonshotai/kimi-k3": { "name": "Kimi K3", "limit": { "context": 1048576, "output": 32768 } }
      }
    }
  },
  "model": "uberllm/z-ai/glm-5.2"
}

{env:UBERLLM_API_KEY} works only in the global file (or a file named by KILO_CONFIG). In a project kilo.jsonc, Kilo ignores it with a warning and the provider has no key. In a project file, paste the key into apiKey instead, or keep the provider in the global file.

Checked on 2026-10-30: each one answered and made a tool call through MartLLM.

Model id Context Max output
deepseek/deepseek-v4.1-flash 1,048,576 384,000
z-ai/glm-5.2 1,048,576 32,768
moonshotai/kimi-k3 1,048,576 32,768

Kilo already sends its app name (X-Title: Kilo Code) to custom providers.

Common errors

What you see Cause and fix
401 / "Invalid API Key" Key missing or mistyped. In a project file, {env:...} is ignored: see above.
404 with "page not found" Base URL must be https://api.martllm.com/v1.
404 model_not_found Model id mistyped. Use Fetch models or copy from the catalog.
402 insufficient_credits Balance empty. Add credit under Dashboard → Credits.
Context fills up at once, or compaction never runs The model has no limit. Add context and output as above.