Route your first request

1. Get an API key

Create an account — no card required. New organizations receive 5,000 free executions and $1.00 of Managed inference credit, both good for 14 days, so the request below works before you connect any provider of your own.

2. Change your base URL

const client = new OpenAI({
  apiKey: process.env.PLANVERITY_API_KEY,
  baseURL: "https://api.planverity.ai/v1"
});

That is the whole integration. Streaming, tool calls and the response shape are unchanged, so an existing client does not know the difference.

3. Send a request

curl https://eicp.onrender.com/v1/chat/completions \
  -H "authorization: Bearer $PLANVERITY_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "model": "auto",
    "messages": [{"role":"user","content":"say hi"}],
    "max_tokens": 64
  }'

model: "auto" lets Planverity pick the lowest-cost model that meets your policy; a specific model or pool works the same way a normal OpenAI request does.

4. Read what happened

Every request writes an execution record before the first provider call — which model answered, why it was selected over the alternatives, what funded it, and what it cost. Open the console to see one.

Bring your own keys, or not

Connect your own Anthropic, OpenAI, Gemini or Z.ai credentials any time under Provider keys in the console, or connect a self-hosted Ollama endpoint. Without one, requests run on prepaid Managed credit — never silently on Planverity’s account under your name. See pricing for exact rates.