Start with LaneZero

What LaneZero is

LaneZero puts one Chat Completions API in front of multiple model providers. Requests use OpenAI-compatible JSON. Model IDs include a provider prefix. Your provider credentials come from your encrypted vault or from a request header.

LaneZero does not sell model credits. Your provider account may still charge for model use.

Zero data retention

Prompt and completion zero data retention: LaneZero does not persist prompt or completion content. It retains only usage metadata—time, model/provider, token counts, estimated cost, latency, status, attempts, and account/key attribution—for reporting.

Quickstart

1. 01

   Sign in with GitHub

   Open the sign-in page. GitHub sign-in creates the browser session used by the management API and console.

2. 02

   Choose how to provide a provider key

   Store a key in the vault from the console. Or keep it out of LaneZero storage and send it with x-provider-key on each request.

3. 03

   Mint a gateway key

   Use the console to create a gateway key. It starts with lz-live_. The full key is shown only once at creation. Copy it before closing the dialog.

4. 04

   Make the first call

   Send the gateway key as a bearer token. Send a provider key too when you chose passthrough mode.

Your first request

Set these values in your shell. Replace the example secrets with your own.

export LZ_URL='http://localhost:8787'
export LZ_KEY='lz-live_REPLACE_ME'
export OPENAI_API_KEY='sk-REPLACE_ME'

Non-streaming chat completion

curl "$LZ_URL/v1/chat/completions" \
  -H "Authorization: Bearer $LZ_KEY" \
  -H "x-provider-key: $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-4.1",
    "messages": [
      {"role": "user", "content": "Write one sentence about pit stops."}
    ]
  }'

Streaming chat completion

curl -N "$LZ_URL/v1/chat/completions" \
  -H "Authorization: Bearer $LZ_KEY" \
  -H "x-provider-key: $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-4.1",
    "messages": [
      {"role": "user", "content": "Count down from three."}
    ],
    "stream": true
  }'

If the OpenAI key is already in your vault, remove the x-provider-key line. The gateway bearer key is always required for /v1 routes.

Where to go next