Skip to main content
Lightweight is 100% OpenAI-compatible. If your tool supports custom OpenAI endpoints, you can connect it to Lightweight in four steps.

​
Generic Setup

1

Find the provider configuration

Open your tool’s settings and look for an OpenAI, OpenAI-compatible, or Custom API provider option.
2

Set the Base URL

Enter the Lightweight API base URL:
https://api.lightweight.one/v1
3

Set the API Key

Enter your Lightweight API key (starts with lw_sk_):
lw_sk_your-key-here
4

Choose a model

Enter any model ID from the catalog. For example:
  • gpt-5.4 — OpenAI flagship
  • claude-sonnet-4.5 — Anthropic balanced
  • gemini-2.5-pro — Google general
  • grok-code-fast-1 — xAI code
If the tool has a model dropdown, you may need to manually type the model name.

​
Environment Variables

Many CLI tools accept configuration via environment variables. These two cover most cases: 
export OPENAI_API_KEY="lw_sk_your-key-here"
export OPENAI_BASE_URL="https://api.lightweight.one/v1"
Some tools use OPENAI_API_BASE instead of OPENAI_BASE_URL. Check your tool’s documentation if the above doesn’t work.

​
Verify It Works

Make a quick test call to confirm the connection: 
curl https://api.lightweight.one/v1/chat/completions \
  -H "Authorization: Bearer lw_sk_your-key-here" \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-5.4", "messages": [{"role": "user", "content": "Hello!"}], "max_tokens": 50}'
If you get a response, your tool will work with the same credentials.
Don’t see your tool listed in our integration guides? Open a discussion and we’ll help you set it up.