Appearance
Getting Started
Universal API is a marketplace and runtime platform for MCP Servers and AI Agents. Build, deploy, and monetize AI tools — or use tools built by others — all through a single API.
What Can You Do?
🔌 Use MCP Servers with Claude & Cline
Connect your AI assistant to powerful tools — Google Calendar, Gmail, Drive, web search, AWS services, document analysis, and more. One config snippet, instant access.
json
{
"mcpServers": {
"universalapi": {
"url": "https://mcp.api.universalapi.co/mcp/mcp-xxx",
"headers": {
"Authorization": "Bearer YOUR_TOKEN"
}
}
}
}🤖 Build & Deploy AI Agents
Create conversational AI agents powered by AWS Strands and AWS Bedrock. Your agents get streaming responses, multi-turn conversations, and access to any MCP server's tools.
💰 Monetize Your Work
Set per-invocation pricing on your resources. Earn revenue when others use your MCP servers or agents. Get paid monthly via Stripe Connect.
Quick Links
| I want to... | Go here |
|---|---|
| Connect Claude/Cline to tools | Quick Start |
| Understand authentication | Authentication |
| Build an MCP Server | Creating MCP Servers |
| Build an AI Agent | Creating Agents |
| See all API endpoints | API Reference |
| Search for resources | Semantic Search |
Platform Architecture
┌─────────────────────────────────────────────────────┐
│ Your AI Client │
│ (Claude Desktop, Cline, curl, SDK) │
└──────────────────────┬──────────────────────────────┘
│ Authorization: Bearer uapi_ut_...
▼
┌─────────────────────────────────────────────────────┐
│ api.universalapi.co │
│ (API Gateway + PublicAuthorizer) │
├──────────────────────┬──────────────────────────────┤
│ MCP Servers │ AI Agents │
│ (Node.js) │ (Python) │
│ │ Strands SDK │
│ Tools │ Streaming │
│ Resources │ Multi-turn │
│ Prompts │ Bedrock LLM │
└──────────────────────┴──────────────────────────────┘How Credits Work
Every API call costs credits. New accounts start with 100 free credits.
- MCP Servers: 1 credit per tool call (minimum)
- AI Agents: 2-20 credits per chat (varies by model and token usage)
Authors can set additional per-invocation pricing. See Pricing for details.