# APIClaw [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://apiclaw.biz/mcp  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/apiclaw

## Description
APIClaw is a flat-rate, OpenAI-compatible AI API gateway exposed as an MCP server. One key and one OpenAI-compatible endpoint give agents access to Claude, GPT, Gemini, DeepSeek, Qwen and more on a flat monthly plan with a daily request quota instead of per-token billing. The MCP endpoint lets clients list models, check quota and issue chat completions. Free tier: 50 requests.

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: medium). Install path inferred — verify against the README before running it. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "apiclaw": {
    "url": "https://apiclaw.biz/mcp"
  }
}
```

## Documentation

## What APIClaw MCP server does

APIClaw exposes an OpenAI-compatible AI API gateway through the Model Context Protocol. Agents connect through one API key and one endpoint, then use the gateway to work with multiple model families, including Claude, GPT, Gemini, DeepSeek, and Qwen. The service is intended for applications that need model access without configuring a separate provider connection for each model family.

Its billing model uses a flat monthly plan with a daily request quota rather than per-token charges. A free tier provides 50 requests, making it suitable for small evaluations or limited agent workflows.

## How it works

An MCP client connects to the APIClaw endpoint and authenticates with the supplied API key. The agent can then invoke MCP operations to discover models, check remaining quota, or submit a chat-completion request. Chat requests use the gateway's OpenAI-compatible API behavior while the agent interacts with the service through MCP.

The model catalog includes Claude, GPT, Gemini, DeepSeek, Qwen, and additional models. Availability is handled through the gateway rather than requiring the agent to connect directly to each underlying provider.

## Setup and configuration

Use the APIClaw MCP endpoint as the connection target and provide an APIClaw key for authentication. The service is designed around a single key and a single OpenAI-compatible endpoint, so applications do not need separate credentials for each listed model family.

After connecting, use the model-listing capability to identify available models before sending chat requests. Quota checks can help an agent monitor its daily request allowance. The free tier includes 50 requests; paid access follows the service's flat monthly plan and daily quota model.

## Tools and capabilities

- List available models through MCP.
- Check the account's request quota.
- Issue chat-completion requests.
- Access Claude, GPT, Gemini, DeepSeek, Qwen, and other models through one gateway.
- Use an OpenAI-compatible API endpoint.

## Limitations and notes

APIClaw uses request quotas rather than unlimited usage: plans include a daily request allowance, and the free tier is limited to 50 requests. The service therefore fits workloads that can operate within a request-based limit. Its main interface is chat completion, alongside model discovery and quota inspection; applications requiring other API operations should verify that those operations are available before adopting it.

_Full upstream README: https://allmcps.com/mcp/apiclaw/readme_