# Clearbar model verdicts

> Clearbar model verdicts is a freemium MCP server for ai models and llm tooling. Look up the cheapest LLM that passes a task, with pass rates and cost per success. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Platforms:** Web, API
- **Website:** https://clearbar.grain64.com/docs/
- **MCP Registry name:** com.grain64.clearbar/verdicts
- **Categories:** AI models and LLM tooling
- **Listing updated:** 2026-11-15

## About

Clearbar prices a model by what a correct answer costs, not by token price. Every attempt, including retries after invalid output or API errors, is summed through OpenRouter's reported cost and divided by the tasks that passed. A model that looks cheap per token but needs three tries shows its real cost.

The MCP server answers from stored verdicts on five public datasets: Amazon Polarity sentiment, glaive function-calling arguments, SMS spam, BANKING77 intents and JSON extraction. Each has a pass-rate bar and a dated OpenRouter price snapshot. Answers include the cheapest passing model, pass rate, cost per successful task, the run file and a script to repeat the run. Re-runs are expected to land within about 5 points of pass rate and 10% of cost.

The server is read-only, needs no account or key and works with any client that supports streamable HTTP.

## Features

- **Cost per successful task:** Retries and failed attempts are priced in and divided by the passes.
- **Stored verdicts:** Covers sentiment, tool calls, spam, banking intents and JSON extraction.
- **Pass bar with intervals:** Statuses use a 95% Wilson interval: cleared, borderline, below bar or errored.
- **Reproducible runs:** Each verdict links a script to repeat it with your own OpenRouter key.
- **Read-only access:** Two tools, no account and no key.

## Good for

- Pick a cheaper model for a spam or intent classifier
- Compare model cost on tool-call argument accuracy
- Ask a coding agent which model to use for JSON extraction
- Re-check a model choice when prices change

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http verdicts https://clearbar.grain64.com/mcp
```

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

```sh
https://clearbar.grain64.com/mcp
```

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

```json
{
  "mcpServers": {
    "verdicts": {
      "url": "https://clearbar.grain64.com/mcp"
    }
  }
}
```

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

```json
{
  "servers": {
    "verdicts": {
      "type": "http",
      "url": "https://clearbar.grain64.com/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "verdicts": {
      "serverUrl": "https://clearbar.grain64.com/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "verdicts": {
      "type": "http",
      "url": "https://clearbar.grain64.com/mcp"
    }
  }
}
```

## Tools

- `list_verdicts`: List every verdict with its cheapest passing model, optionally filtered by task
- `get_verdict`: Return the full table for one task by slug

## FAQ

### Which tools does it expose?

list_verdicts, with an optional task filter, and get_verdict, which takes a slug.

### Do I need a key?

No. The MCP server is open and read-only. Running your own verdict uses your own OpenRouter key.

### Can I test my own examples?

Yes, in the Clearbar workspace: 3 free runs a month with up to 100 examples and 6 models.

### How repeatable are the numbers?

Models run at temperature 0 but are not fully deterministic, so expect about 5 points of pass rate and 10% of cost.

### Is Clearbar model verdicts free?

Clearbar model verdicts has a free plan, with paid plans for more usage or features.

### How do I add Clearbar model verdicts to Claude Code?

Run this command in your terminal: claude mcp add --transport http verdicts https://clearbar.grain64.com/mcp

### What are the best alternatives to Clearbar model verdicts?

The closest alternatives to Clearbar model verdicts 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/clearbar-model-verdicts
