# ModelSpec

> ModelSpec is a freemium MCP server for ai models and llm tooling. Shortlist AI models for a job from requirements, sourced benchmarks and real cost. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Starting price:** $29 a month
- **Platforms:** API, CLI
- **Website:** https://modelspec.dev
- **Repository:** https://github.com/turbobeest/modelspec/tree/HEAD/mcp
- **MCP Registry name:** io.github.turbobeest/modelspec
- **Categories:** AI models and LLM tooling
- **Listing updated:** 2026-11-15

## About

ModelSpec frames model choice as an analysis of alternatives rather than a routing call. Hard requirements act as gates, so a model that fails one stays on the board marked excluded with the reason, while weighted preferences only reorder the models that passed. Each answer also carries uncertainty: when two models cannot be separated on the evidence, both are returned as a tie. The site's example is coding, where Claude Sonnet 5.5 has a 39% chance of scoring at least as well as Claude Opus 5.5, so both are listed, at $1,200 and $2,400 a month for 10,000 tasks on its 14 November 2026 price snapshot.

The MCP server is a thin read-only proxy, stateless, running on Cloudflare Workers. Each tool call hits the public rank or policy-check API or a static JSON export and returns the body with its origin URL. Data tools need an API key sent as a bearer token or `x-api-key`. A public server card needs no key.

The board covers 44 decidable models out of 1,372 catalogued cards. The engine is MIT licensed and the data CC BY-SA. The company says it takes no referral fees or paid placement, and answers can be reproduced from a signed snapshot, spec hash and decision ID.

## Features

- **Requirement gates:** Must conditions exclude models, which remain visible with the reason.
- **Tie reporting:** Returns a tie when the evidence cannot separate the top models.
- **Cost per task:** Estimates cost at your token counts and plan break-even.
- **Licence and residency check:** policy_check returns cited commercial-use and data-residency determinations.
- **Audit trail:** A signed snapshot, spec hash and decision ID let an answer be reproduced.
- **Independence commitment:** No referral fees or paid placement, per the published commitment.

## Good for

- Pick a model for a coding agent based on evidence and monthly cost
- Check whether a model's licence allows commercial use
- Decide which models belong on a router's allow-list
- Compare two models for a given capability
- Keep a record of why a model was chosen

## Install

### Claude Code (Terminal)

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

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

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

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

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

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

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

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

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

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

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

## Tools

- `rank`: Rank models for one capability.
- `model_info`: Read the card for one model.
- `list_use_cases`: List the use-case profiles.
- `policy_check`: Check licence and data-residency determinations.
- `decide`: Return a decision on which models fit a spec.
- `compare_models`: Compare models against each other.

## FAQ

### Is it free?

People can use the board on the site for free. MCP, API and CLI access uses prepaid credits and an API key.

### What do answers cost?

A decision costs 1 credit, a fully explained decision 2, and a licence and residency check 5; errors and no-fit answers are free.

### Is it a router?

No. It sits outside the request path and answers which models are worth routing to; prompts are not kept.

### Which plans exist?

Solo $29/mo for 2,500 credits, Team $199/mo for 25,000 and Scale $799/mo for 150,000, plus credit packs from $5.

### Is ModelSpec free?

ModelSpec has a free plan, with paid plans for more usage or features. Paid plans start at $29 a month.

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

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

### What are the best alternatives to ModelSpec?

The closest alternatives to ModelSpec 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/modelspec
