# Unprofane

> Unprofane is a freemium MCP server for security. Profanity and toxicity checks that see through disguised text, with evidence per verdict. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Starting price:** $9 a month
- **Platforms:** API
- **Website:** https://unprofane.com/docs/mcp/
- **MCP Registry name:** com.unprofane/unprofane
- **Categories:** security, developer tools
- **Listing updated:** 2026-11-15

## About

Five tools wrap the same engine and keys as the Unprofane REST API. `check_text` takes text plus context, policy, mode, language, country and mask settings. `check_batch` accepts up to 100 items and 30,000 characters per call. `report_result` sends a correction for a result id, `get_usage` returns checks used and remaining, and `get_integration_snippet` returns a copy-paste recipe for a platform such as Supabase, Clerk, Firebase, Unity or Godot in a chosen language.

Each tool declares readOnlyHint and destructiveHint. All are read-only except `report_result`, which writes a report and destroys nothing, so a client can run checks without prompting. Results come back as structured content against an output schema. Only checked texts count against your allowance; the other tools count nothing.

The pricing page says the filter catches 13 kinds of disguise across 17 languages and shows the words found, where they were and a censored copy. It also claims 89.0 percent of toxic texts caught versus 58.4 percent for Perspective at the same false-alarm rate, a figure the vendor publishes with its own test. The server is stateless: every message is a POST and GET or DELETE return 405. Publishable keys are refused; use a bearer secret key.

## Features

- **Single text check:** check_text returns a verdict with context, policy and language settings.
- **Batch checks:** check_batch takes up to 100 items and 30,000 characters in one call.
- **Correction reports:** report_result sends feedback on a result by its id.
- **Usage readout:** get_usage shows checks used, included and over for the period.
- **Integration recipes:** get_integration_snippet returns setup code for platforms like Supabase, Clerk, Unity and Godot.
- **Annotated tools:** Every tool carries read-only and destructive hints and a typed output schema.

## Good for

- Moderate a queue of user comments inside an agent loop
- Let a coding agent add a profanity filter to a game or web app
- Check usernames and chat messages before publishing
- Report a missed or wrong verdict from the same assistant

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http unprofane https://api.unprofane.com/mcp
```

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

```sh
https://api.unprofane.com/mcp
```

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

```json
{
  "mcpServers": {
    "unprofane": {
      "url": "https://api.unprofane.com/mcp"
    }
  }
}
```

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

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

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

```json
{
  "mcpServers": {
    "unprofane": {
      "serverUrl": "https://api.unprofane.com/mcp"
    }
  }
}
```

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

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

## Tools

- `check_text`: Checks one text for profanity and toxicity.
- `check_batch`: Checks up to 100 items in one call.
- `report_result`: Sends a correction for a result by its id.
- `get_usage`: Returns this period's checks used, included and over.
- `get_integration_snippet`: Returns a copy-paste integration recipe for a platform and language.

## FAQ

### How do I authenticate?

Send a bearer secret key, the same as the /v1 API; publishable keys are refused.

### What does it cost?

Plans run from Hobby at $9 a month for 25,000 checks to Scale at $999 a month for 5,000,000 checks. Free test keys give 100 checks, and Hobby and Starter are half price for the first three months until December 31, 2026.

### What counts as a check?

Each text checked counts as one; the other tools count nothing.

### How many languages?

17 languages, according to the pricing page.

### Does the server keep sessions?

No. Every message is a POST and GET or DELETE return 405.

### Is Unprofane free?

Unprofane has a free plan, with paid plans for more usage or features. Paid plans start at $9 a month.

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

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

### What are the best alternatives to Unprofane?

The closest alternatives to Unprofane 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/unprofane
