# Epoch Ai Models

> Epoch Ai Models is a freemium MCP server for ai models and llm tooling. MCP server giving AI clients lookups into Epoch AI's database of notable AI models. It can run locally on your machine or as a hosted remote server.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Starting price:** $49 a month
- **Platforms:** API, CLI
- **Website:** https://pipeworx.io/packs/epoch-ai-models
- **Repository:** https://github.com/pipeworx-io/mcp-epoch-ai-models
- **MCP Registry name:** io.github.pipeworx-io/epoch-ai-models
- **Categories:** [AI models and LLM tooling](https://ai-powered.com/categories/ai-and-llm)
- **Listing updated:** 2026-11-18

## About

The README describes this server as a way to pull facts about notable machine learning models from Epoch AI's dataset: organization, release date, parameter count, training-compute estimate, hardware and accessibility. A lookup by name resolves dated variants, so a bare model name matches the canonical dated release in the dataset. Filtered searches cover organization, domain, publication year and a minimum training-compute threshold, and a yearly compute trend tool summarises the largest training run per year.

The server is hosted on the Pipeworx gateway and connects over streamable HTTP or through an npm stdio bridge. Calls are keyless and anonymous by default, with a daily cap that a free Pipeworx account raises. Each tool response carries Epoch AI's CC BY attribution, according to the README.

It suits researchers, analysts and developers who want an agent to answer questions about model scale or training compute from a structured dataset rather than from web search. The README notes that many entries lack compute or parameter estimates, so those fields can be empty.

## Features

- **Single model lookup:** Returns organization, release date, parameters, compute estimate, hardware and accessibility for one named model.
- **Dated variant matching:** Resolves a bare model name such as GPT-4 to its dated release in the dataset.
- **Filtered model search:** Filters by organization, domain, publication year and minimum training compute, with paginated results.
- **Yearly compute trend:** Summarises per-year model counts and the largest training run, optionally by domain and start year.
- **Keyless anonymous access:** Works without an account at a capped daily rate, with an optional Pipeworx key for a higher limit.

## Good for

- Comparing training compute and parameter counts across model releases
- Finding models from a given organization or domain released in a specific year
- Tracking how the largest training runs have grown year over year
- Giving an agent a structured source for AI model history questions

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http epoch-ai-models https://gateway.pipeworx.io/epoch-ai-models/mcp
```

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

```sh
https://gateway.pipeworx.io/epoch-ai-models/mcp
```

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

```json
{
  "mcpServers": {
    "epoch-ai-models": {
      "url": "https://gateway.pipeworx.io/epoch-ai-models/mcp"
    }
  }
}
```

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

```json
{
  "servers": {
    "epoch-ai-models": {
      "type": "http",
      "url": "https://gateway.pipeworx.io/epoch-ai-models/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "epoch-ai-models": {
      "serverUrl": "https://gateway.pipeworx.io/epoch-ai-models/mcp"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "epoch-ai-models": {
      "type": "http",
      "url": "https://gateway.pipeworx.io/epoch-ai-models/mcp"
    }
  }
}
```

## Tools

- `ai_model`: Looks up one AI model's organization, release date, parameter count, training-compute estimate, hardware and accessibility.
- `ai_models_search`: Filters the models dataset by organization, domain, year and minimum training compute.
- `ai_compute_trend`: Returns per-year model counts and the largest training run by year, optionally filtered.

## FAQ

### Does the server require an API key?

No. The README says it is keyless, and anonymous calls are capped at 50 a day. A free Pipeworx account raises the cap to 200 a day.

### What happens when a model name is not found?

The README says an unmatched name returns found:false along with the closest names in the dataset rather than an empty result.

### Are all models in the dataset given compute estimates?

No. The README notes that many entries leave training-compute and parameter fields blank, and those rows are excluded from compute filters and trends rather than treated as zero.

### Which transports does the server support?

The registry entry lists stdio via the npm package @pipeworx/mcp-epoch-ai-models and streamable HTTP at the gateway endpoint.

### Is Epoch Ai Models free?

Epoch Ai Models has a free plan, with paid plans for more usage or features. Paid plans start at $49 a month.

### How do I add Epoch Ai Models to Claude Code?

Run this command in your terminal: claude mcp add --transport http epoch-ai-models https://gateway.pipeworx.io/epoch-ai-models/mcp

### What are the best alternatives to Epoch Ai Models?

The closest alternatives to Epoch Ai Models are AI Model Analyzer, Agentic Weekly, The Neural Net, MCP and Fireworks AI MCP by usefulapi.

## Where this comes from

- Description: written from the project's README, the project's website and its MCP Registry entry.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [AI Model Analyzer](https://ai-powered.com/mcp/ai-model-analyzer): AI Model Analyzer is a free MCP server for ai models and llm tooling.
- [Agentic Weekly](https://ai-powered.com/mcp/agentic-weekly): Agentic Weekly is a free MCP server for ai models and llm tooling.
- [The Neural Net](https://ai-powered.com/mcp/the-neural-net): The Neural Net is a free MCP server for ai models and llm tooling.
- [MCP](https://ai-powered.com/mcp/mcp-mwi57hue): MCP is a paid MCP server for design and media.
- [Fireworks AI MCP by usefulapi](https://ai-powered.com/mcp/fireworks-ai-mcp-by-usefulapi): Fireworks AI MCP by usefulapi is a freemium MCP server for ai models and llm tooling.
- [Together AI MCP by usefulapi](https://ai-powered.com/mcp/together-ai-mcp-by-usefulapi): Together AI MCP by usefulapi is a freemium MCP server for ai models and llm tooling.
- [Replicate](https://ai-powered.com/mcp/replicate): Replicate is a freemium MCP server for ai models and llm tooling.
- [querying.ai](https://ai-powered.com/mcp/querying-ai): querying.ai is a freemium MCP server for ai models and llm tooling.

---
Source: https://ai-powered.com/mcp/epoch-ai-models
