Calypso

Calypso RAG

Product docs

Quick Start |
Getting startedOpenAI Compatible APIBucketsIntegrationsReference
Getting Started

Quick start

Get a project API key, prepare a knowledge bucket, call the default Calypso RAG agent, and choose the right deployment path.

Get a project API key

To call Calypso RAG from an API client, MCP server, or automation, you need an active project API key.

You can get one by:

  • signing in to your Calypso RAG workspace or joining for free
  • opening Project → API Keys
  • creating a new key or revealing an existing active key
Treat project API keys like secrets. Do not commit them to git or expose them in untrusted clients.

Prepare a bucket-backed source

Before your first API request, add at least one source to a knowledge bucket in your workspace.

You can do that either:

Start with something small and trusted:

  • a file with durable product or policy information
  • or a website that acts as a canonical public source

Every durable upload needs a bucket destination. Use a small, trusted bucket first, then wait until indexing and bucket sync are ready for retrieval.

If needed, confirm in Agent Builder that the default model is:

calypso-rag-agent

and that its retrieval scope includes the bucket you want to test. Named profiles use model ids such as calypso-rag-agent:{profile_id} and can point at different buckets.

Make your first API call

Here is how to send your first grounded request with the default Calypso RAG agent. The project API key determines which workspace, buckets, and named agent variants are available.

Python

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_CALYPSO_API_KEY",
    base_url="https://api.calypso.so/v1"
)

response = client.responses.create(
    model="calypso-rag-agent",
    input="Summarize the grounded knowledge available in this workspace."
)

print(response.output_text)

JavaScript

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "YOUR_CALYPSO_API_KEY",
  baseURL: "https://api.calypso.so/v1",
});

const response = await client.responses.create({
  model: "calypso-rag-agent",
  input: "Summarize the grounded knowledge available in this workspace.",
});

console.log(response.output_text);

cURL

curl -X POST "https://api.calypso.so/v1/responses" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_CALYPSO_API_KEY" \
  -d '{
    "model": "calypso-rag-agent",
    "input": "Summarize the grounded knowledge available in this workspace."
  }'

To call a named profile, set model to the variant id, for example calypso-rag-agent:support.

After the call succeeds, validate the same request in Playground so you can inspect grounded behavior, confirm the right project key is selected, and verify the answer is using the bucket-backed source you loaded.

Next steps

Support

If your first request does not behave the way you expect:

Next page

Project API keys