---
title: "Multimodal Gemini File Search RAG for Websites & AI Agents | Calypso"
canonical_url: "https://www.calypso.so/"
last_updated: "2026-06-25T02:21:46.977Z"
meta:
  description: "Build multimodal Gemini File Search RAG for websites and AI agents. Answer from PDFs, docs, images, diagrams, and help content with citations."
  keywords: "multimodal RAG, Gemini File Search RAG, Google Gemini File Search, website RAG, AI answer widget, RAG with citations, RAG for PDFs and images, chat with PDFs and images, source-backed answers, RAG knowledge buckets, bucket-backed uploads, multimodal knowledge base, AI agent knowledge base, Calypso RAG"
  "og:description": "Build multimodal Gemini File Search RAG for websites and AI agents. Answer from PDFs, docs, images, diagrams, and help content with citations."
  "og:title": "Multimodal Gemini File Search RAG for Websites & AI Agents | Calypso"
  "twitter:description": "Build multimodal Gemini File Search RAG for websites and AI agents. Answer from PDFs, docs, images, diagrams, and help content with citations."
  "twitter:title": "Multimodal Gemini File Search RAG for Websites & AI Agents | Calypso"
---

Calypso home

Gemini File Search-powered multimodal RAG

# Fully managedRAG for sourcebacked AI answers

Turn PDFs, docs, screenshots, charts, diagrams, help content, and images into reusable source memory for agents — with citations users can verify.

[**Start free **](https://rag.calypso.so/join) [**Read docs **](https://docs.calypso.so/)

Bucket-backed retrieval, harnessed agents, and citations — without maintaining your own RAG pipeline.

**Quick start **

**OpenAI-compatible Calypso Agent **

```
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["CALYPSO_API_KEY"],
    base_url="https://api.calypso.so/v1",
)

response = client.responses.create(
    model="calypso-rag-agent:spacex",
    input="Summarize SpaceX's business model and cite sources.",
    reasoning={"effort": "low"},
)

print(response.output_text)
```

**CALYPSO_API_KEY ****Python **

[**MCP Server****@calypsohq/multimodal-rag-mcp-server** Connect Calypso RAG to Cursor, Claude Desktop, and MCP-compatible AI clients.](https://www.npmjs.com/package/@calypsohq/multimodal-rag-mcp-server) [**n8n Node****@calypsohq/n8n-nodes-calypso** Ask agents, load named profiles, inspect buckets, and upload files from n8n workflows.](https://www.npmjs.com/package/@calypsohq/n8n-nodes-calypso) [**Web Widget****Website answer surface** Embed a polished source-backed chat widget on your site with the same hosted agent layer.](https://www.calypso.so/integrations) [**API Endpoint****OpenAI-compatible agent calls** Call hosted Calypso Agents from OpenAI-style clients and agent frameworks without rebuilding retrieval.](https://www.calypso.so/agents)

**Product architecture **

## One source layer. Harnessed agents. Answers everywhere.

Buckets hold multimodal unstructured knowledge. Agents add scope, behavior, and citation policy. Integrations bring the same grounded agent to your website, API, workflows, and AI clients.

BucketsAgentsIntegrations

**01 / Calypso Buckets**

### The source layer agents can trust.

Buckets give every upload an owner, scope, and destination before retrieval begins.

**Layer 01**

**Source layer**

#### **Organize multimodal sources**

Upload PDFs, docs, websites, screenshots, and Q&A into named Buckets agents can search.

**01**

**Upload routing**

#### **Route knowledge with intent**

Tie every source to a bucket id, readable slug, and retrieval scope before indexing.

**02**

**Trust signal**

#### **Know what is ready to cite**

Track accepted, indexed, and active states so agents only search ready knowledge.

**03**

**02 / Calypso Agents**

### A policy layer for grounded behavior.

Agents decide where to search, how much effort to spend, and how evidence is presented.

**Layer 02**

**Hosted runtime**

#### **Start with a hosted agent**

Use calypso-rag-agent, then create named profiles for each team or surface.

**01**

**Curated memory**

#### **Attach the right Buckets**

Scope each agent to approved source sets so retrieval stays intentional.

**02**

**Verifiable output**

#### **Return evidence by default**

Set citation behavior, response effort, and answer policy once, then reuse it.

**03**

**03 / Calypso Integrations**

### One RAG layer for every interface.

Integrations turn the same hosted knowledge layer into customer, workflow, and developer surfaces.

**Layer 03**

**Website**

#### **Drop answers onto your site**

Launch a polished widget that answers from selected Buckets and shows sources.

**01**

**API**

#### **Call Agents like a model**

Use OpenAI-compatible requests with bucket memory and citation-ready responses.

**02**

**MCP + workflows**

#### **Move knowledge into RAG**

Connect MCP clients and n8n workflows to inspect Buckets, upload files, and call Agents.

**03**

**Pricing **

## Hosted RAG plans that feel simple.

Pay for answers, not tokens. Starter and Pro begin with a 14-day trial using Pro entitlements, then stay predictable with fixed monthly allowances and prepaid Answer Packs.

Starter and Pro include a 14-day trial with Pro entitlements. Upgrade only when the answer layer is working for your team.

**Plan**

### **Starter**

$19/month

Launch a grounded answer bot with no surprise bills.

1,000 answers / month

100 MB knowledge storage

1 agent profile

All integrations: API keys, MCP, n8n, web widget

[**Start 14-day free trial**](https://rag.calypso.so/join?targetAppKey=rag&returnTo=%2Fcheckout%3FragPlan%3Dstarter%26source%3Donboarding)

14-day trial with Pro entitlements

**Most common**

### **Pro**

$49/month

Everything you need to ship RAG into production.

5,000 answers / month

1 GB knowledge storage

10 agent profiles

All integrations: API keys, MCP, n8n, web widget

Insights analytics included

[**Start 14-day free trial**](https://rag.calypso.so/join?targetAppKey=rag&returnTo=%2Fcheckout%3FragPlan%3Dpro%26source%3Donboarding)

14-day trial with Pro entitlements

**Plan**

### **Scale**

$199/month

High-volume RAG with priority throughput.

25,000 answers / month

10 GB knowledge storage

Unlimited agent profiles

All integrations + Insights analytics

Priority throughput

Custom terms available — talk to us

[**Start 14-day free trial**](https://rag.calypso.so/join?targetAppKey=rag&returnTo=%2Fcheckout%3FragPlan%3Dultra%26source%3Donboarding)

14-day trial with Pro entitlements

**Pricing guarantee **

**No surprise bills.** Every plan includes a fixed monthly answer allowance. Need more capacity? Add a prepaid Answer Pack — 1,000 extra answers for $10. No metered overage. Every plan includes 1 seat. Add teammates anytime for $19 / seat / month — you’re only billed for the seats you use.

**Questions and answers **

## Questions before shipping grounded answers.

Most questions come down to retrieval, source coverage, citations, and whether the same answer layer can be reused beyond the website.

**Launch Calypso RAG **

## **Put grounded answers wherever work happens. **

Use the same Bucket-backed agents across your website, API, MCP clients, n8n workflows, and product UI — with citations users can verify.

[**Start free **](https://rag.calypso.so/join)