# RealRun

> RealRun is a free MCP server for ai models and llm tooling. Independent AI model scores on inbox, scheduling, receipts, finance and support replies. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Free
- **Platforms:** API, CLI
- **Website:** https://realrun.baseledgerip.com
- **Repository:** https://github.com/baseledgerip/realrun-mcp
- **MCP Registry name:** com.baseledgerip.realrun/realrun
- **Categories:** AI models and LLM tooling
- **Listing updated:** 2026-11-15

## About

Choosing a model for a routine business task is easier with test results than with a hunch. RealRun publishes its own scores for AI models on five everyday jobs, and this server makes those headline numbers available to an agent so it can ask which model to use for a given task.

The five task ids are inbox-triage, meeting-scheduling, receipts, finance and support-replies. The site says each test uses realistic items, fresh held-back sets the prompts were never tuned on, scam-safety checks and published methods, with 9 to 10 models per test. The recommend_model tool picks one model plus two alternatives by accuracy, cost or a balance, and never picks a model flagged as not recommended.

All tools are read-only and stateless. Full prompts and item-level results sit in paid playbooks ($29 each, or $99 for all five). The README says metering is planned but switched off, so nobody is charged for lookups.

## Features

- **Per-task scores:** Returns every tested model's headline score for a task, best first.
- **Model recommendation:** Suggests one model and two alternatives by accuracy, cost or balance.
- **Warning flags:** Lists not-recommended and caution flags such as scam failures or invented data.
- **Task catalogue:** Lists tested tasks, item counts, dates and playbook prices.
- **Read-only and stateless:** Stores nothing and needs no authentication.

## Good for

- Choose a model for email triage
- Check which models fell for scam invoices
- Pick the cheapest model that scores well on receipts
- Add a model-choice step to an agent workflow

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http realrun https://realrun.baseledgerip.com/mcp
```

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

```sh
https://realrun.baseledgerip.com/mcp
```

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

```json
{
  "mcpServers": {
    "realrun": {
      "url": "https://realrun.baseledgerip.com/mcp"
    }
  }
}
```

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

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

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

```json
{
  "mcpServers": {
    "realrun": {
      "serverUrl": "https://realrun.baseledgerip.com/mcp"
    }
  }
}
```

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

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

## Tools

- `list_tasks`: Tested tasks, item counts, dates, results pages and playbook prices.
- `get_scores`: Every model's headline score for a task, best first, with cost tier.
- `recommend_model`: One pick plus two alternatives for a task by priority.
- `get_warnings`: Not-recommended and caution flags for a task.

## FAQ

### Does it need an account?

No. The README says the server has no auth and stores nothing.

### Is it free?

Lookups are free for now; the README says per-call billing is planned but off. Playbooks cost $29 each or $99 for five.

### Which tasks are covered?

Inbox triage, meeting scheduling, receipts, finance and customer support replies.

### Can I run it locally?

Yes, over stdio with npx -y realrun-mcp, or use the remote endpoint.

### Is RealRun free?

Yes. RealRun is free to use.

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

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

### What are the best alternatives to RealRun?

The closest alternatives to RealRun are Claude Code Ultimate Guide, signafAIde, Toolkien, Lugh and HypePaper.

## Alternatives

- [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.
- [signafAIde](https://ai-powered.com/mcp/signafaide): signafAIde is a freemium MCP server for security.
- [Toolkien](https://ai-powered.com/mcp/toolkien): Toolkien is an open-source MCP server for ai models and llm tooling.
- [Lugh](https://ai-powered.com/mcp/lugh): Lugh is an open-source MCP server for ai models and llm tooling.
- [HypePaper](https://ai-powered.com/mcp/hypepaper): HypePaper is a freemium MCP server for ai models and llm tooling.
- [Agent Data Tools: Compact (find, describe, run)](https://ai-powered.com/mcp/agent-data-tools-compact-find-describe-run): Agent Data Tools: Compact (find, describe, run) is a freemium MCP server for developer tools.
- [Angel Agent Guard: pre-purchase check + prompt-injection scan](https://ai-powered.com/mcp/angel-agent-guard-pre-purchase-check-prompt-injection-scan): Angel Agent Guard: pre-purchase check + prompt-injection scan is a freemium MCP server for ai models and llm tooling.
- [Mizan Calibration](https://ai-powered.com/mcp/mizan-calibration): Mizan Calibration is a free MCP server for ai models and llm tooling.

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