# SuperCompress

> SuperCompress is a freemium MCP server for ai models and llm tooling. MCP and proxy that compresses coding-agent context before it reaches the model. It can run locally on your machine or as a hosted remote server.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Platforms:** CLI, API
- **Website:** https://www.supercompress.dev
- **Repository:** https://github.com/supercompress/supercompress/tree/HEAD/packages/proxy
- **MCP Registry name:** io.github.Supercompress/supercompress
- **Categories:** AI models and LLM tooling, developer tools
- **Listing updated:** 2026-11-16

## About

SuperCompress sits between an agent and the model and scores bulky context, such as logs, diffs, tool traces and pasted files, against the current question. Lines that carry the needed evidence are kept in their original wording and the rest is dropped, rather than rewritten as a summary. The user's own question is never altered.

The vendor reports a 64.1% mean context cut on its coding-agent benchmark, 16,647 tokens down to 5,148, with required evidence kept in 24 of 24 cases, and notes this measures evidence containment, not downstream answer quality or a total bill reduction. The npm package supercompress-proxy installs MCP and hooks across 40+ coding-agent harnesses, including Cursor, Claude Code, Codex and Windsurf, with a browser sign-in. Four MCP tools handle compression, account linking, usage and rate-limit runway.

There is also an API and SDK for chatbots, RAG and support agents. An optional local proxy rewrites an OpenAI-compatible base URL.

## Features

- **Query-aware compression:** Scores context against the current question and drops the rest.
- **Original wording kept:** Keeps evidence lines verbatim instead of summarising.
- **40+ agent harnesses:** One setup command adds MCP and hooks to supported agents.
- **Usage reporting:** Shows plan, quota and savings.
- **Rate-limit runway:** Checks provider rate-limit windows locally.
- **Optional local proxy:** Can rewrite an OpenAI-compatible base URL to localhost.

## Good for

- Shrink a huge log or diff before an agent reads it
- Cut input tokens in a long coding session
- Check how much of a rate-limit window remains
- Compress retrieved documents before a model call

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http supercompress https://api.supercompress.dev/api/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://api.supercompress.dev/api/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "supercompress": {
      "url": "https://api.supercompress.dev/api/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "supercompress": {
      "type": "http",
      "url": "https://api.supercompress.dev/api/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "supercompress": {
      "serverUrl": "https://api.supercompress.dev/api/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "supercompress": {
      "type": "http",
      "url": "https://api.supercompress.dev/api/mcp"
    }
  }
}
```

## Tools

- `compress_context`: Compresses bulky context for a query
- `connect_account`: Links the install to your dashboard
- `usage_summary`: Shows savings for the connected account
- `runway_check`: Checks provider rate-limit runway locally

## FAQ

### What does it cost?

The site lists 1M free tokens a month, then $0.10 per 1M.

### Does it keep the answer-critical lines?

The vendor reports required evidence held in 24 of 24 benchmark cases, measured as evidence containment.

### How do I install it?

npm install -g supercompress-proxy, then run supercompress setup.

### Does it need an API key?

Not necessarily; the setup flow uses a secure browser sign-in and an API key is optional.

### Is SuperCompress free?

SuperCompress has a free plan, with paid plans for more usage or features.

### How do I add SuperCompress to Claude Code?

Run this command in your terminal: claude mcp add --transport http supercompress https://api.supercompress.dev/api/mcp

### What are the best alternatives to SuperCompress?

The closest alternatives to SuperCompress are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Ouroboros](https://ai-powered.com/mcp/ouroboros): Ouroboros is an open-source MCP server for developer tools.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Concord MCP](https://ai-powered.com/mcp/concord-mcp): Concord MCP is a freemium MCP server for developer tools.

---
Source: https://ai-powered.com/mcp/supercompress
