# Lantern Works GPU Prices MCP server

> Compare GPU cloud rental prices, track price changes and estimate GPU job and agent stack costs across tracked providers.

- Listing: https://mcp.tc/i/lantern-works-gpu-prices
- Connect: use the server's own URL `https://lantern-works-mcp.lantern-works.workers.dev/mcp` (no sign-in); clients connect to it directly. The listing link is a page, not an MCP endpoint.
- Type: remote (Streamable HTTP)
- Auth: none
- Category: [Cloud & DevOps](https://mcp.tc/c/cloud-devops)
- Vendor: Lantern Works

## About

Connects to the Lantern Works price feed for GPU cloud rentals. It compares H100, H200, A100, L40S, RTX 4090/5090 and other models across tracked providers, cheapest first, with each price's vendor source link and check time. It also reports price changes and history, estimates GPU job costs, optimizes a monthly stack of compute, scraping, proxy, search and email, and searches a feed of agent work-stack deals.

Runs as a hosted Streamable HTTP endpoint that needs no sign-in or API key, or locally with uvx lantern-works-deals-mcp. Most tools are read-only. The hosted endpoint also offers optional webhook price alerts that return an API key for managing watches.

## What it can do

- Compare GPU rental prices by model, cheapest first
- List recent GPU price changes with sources
- Estimate the cost of a single GPU job
- Optimize a monthly stack cost from needs or a description
- Search and compare agent work-stack deals
- Set webhook alerts for GPU price thresholds

## Tools (13)

- `compare_gpu_prices`: Compare tracked GPU rental prices for a model, cheapest first, with sources and check times. (read-only)
- `get_price_changes`: List GPU price changes with old and new prices, direction and source. (read-only)
- `optimize_stack`: Find the cheapest suitable stack for monthly scraping, proxy, email, search and GPU needs. (read-only)
- `optimize_agent_stack`: Estimate monthly stack cost from a plain-text description of an agent. (read-only)
- `estimate_job_cost`: Estimate the cost of one GPU job across tracked providers. (read-only)
- `rebuild_quote`: Reopen a past quote from its receipt and revalidate against the current feed. (read-only)
- `search_deals`: Search agent work-stack deals by category, offer type and price. (read-only)
- `compare_deals`: Compare 2 to 5 deals side by side. (read-only)
- `get_deal`: Get one deal's public record by id. (read-only)
- `create_watch_key`: Register a webhook and get an API key for price alerts.
- `watch_price`: Create a webhook alert when a GPU rate crosses a price threshold.
- `list_watches`: List your watches and alert counts. (read-only)
- `delete_watch`: Delete a watch by id.

## Example prompts

- "What is the cheapest H100 rental right now?"
- "Show GPU price drops since 2026-11-01"
- "Estimate the cost of a 20-hour job on 4 A100 GPUs"
- "Find free-tier deals for search APIs"

## Install

### Claude Code

1. Run this in a terminal, in your project folder:

```bash
claude mcp add --transport http lantern-works-gpu-prices https://lantern-works-mcp.lantern-works.workers.dev/mcp
```

2. Start Claude Code and type `/mcp`. **lantern-works-gpu-prices** should show as connected.

Add `--scope user` to make it available in every project, not just this one.

### Claude Desktop

1. Open **Customize → Connectors**, click **+ Add**, then **Add custom connector**.

2. Name it **Lantern Works GPU Prices** and paste this URL:

```url
https://lantern-works-mcp.lantern-works.workers.dev/mcp
```

3. Click **Add**. Its tools appear in the chat’s tools menu.

Claude Desktop’s JSON config file only starts local servers. Remote servers go through Connectors, and connectors you add on claude.ai show up here too.

### claude.ai

1. Open the connector form on claude.ai. This button fills in the name and URL for you:

[Add to claude.ai](<https://claude.ai/customize/connectors?modal=add-custom-connector&connectorName=Lantern%20Works%20GPU%20Prices&connectorUrl=https%3A%2F%2Flantern-works-mcp.lantern-works.workers.dev%2Fmcp>) (opens connector settings)

2. Check that the URL reads `https://lantern-works-mcp.lantern-works.workers.dev/mcp` and click **Add**.

3. Turn it on in a chat from the tools menu.

Free plans allow one custom connector. On Team and Enterprise plans an owner adds it under **Organization settings → Connectors**.

### ChatGPT

1. On chatgpt.com, go to **Plugins**, select **+**, then **Add custom MCP server**.

2. Enter a name for the server.

3. Under **Connection**, paste `https://lantern-works-mcp.lantern-works.workers.dev/mcp` as the server URL, then for authentication choose **No authentication**.

4. Check **I understand and want to continue**, then select **Create as a plugin**.

Whether you can add a custom MCP server depends on your ChatGPT plan and your workspace permissions (in Business workspaces, admins and owners can).

### Cursor

[Add to Cursor](<https://cursor.com/install-mcp?name=lantern-works-gpu-prices&config=eyJ1cmwiOiJodHRwczovL2xhbnRlcm4td29ya3MtbWNwLmxhbnRlcm4td29ya3Mud29ya2Vycy5kZXYvbWNwIn0%3D>) (opens Cursor)

Or add it by hand to `~/.cursor/mcp.json` (all projects) or `.cursor/mcp.json` (this project):

`mcp.json`:

```json
{
  "mcpServers": {
    "lantern-works-gpu-prices": {
      "url": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
    }
  }
}
```

### VS Code

[Install in VS Code](<https://vscode.dev/redirect/mcp/install?name=lantern-works-gpu-prices&config=%7B%22type%22%3A%22http%22%2C%22url%22%3A%22https%3A%2F%2Flantern-works-mcp.lantern-works.workers.dev%2Fmcp%22%7D>) (opens VS Code)

Or from a terminal:

```bash
code --add-mcp '{"name":"lantern-works-gpu-prices","type":"http","url":"https://lantern-works-mcp.lantern-works.workers.dev/mcp"}'
```

Or commit it to the repo in `.vscode/mcp.json`:

`.vscode/mcp.json`:

```json
{
  "servers": {
    "lantern-works-gpu-prices": {
      "type": "http",
      "url": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
    }
  }
}
```

### Devin Desktop

1. Add it to `~/.config/devin/mcp_config.json` (macOS and Linux) or `%APPDATA%\devin\mcp_config.json` (Windows):

`mcp_config.json`:

```json
{
  "mcpServers": {
    "lantern-works-gpu-prices": {
      "serverUrl": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
    }
  }
}
```

2. Refresh the MCP server list in Cascade.

Devin Desktop is the new name for Windsurf. It reads `serverUrl` (or `url`) for remote servers.

### Codex

```bash
codex mcp add lantern-works-gpu-prices --url https://lantern-works-mcp.lantern-works.workers.dev/mcp
```

Or edit `~/.codex/config.toml` directly:

`config.toml`:

```toml
[mcp_servers.lantern-works-gpu-prices]
url = "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
```

### Gemini CLI

```bash
gemini mcp add --transport http lantern-works-gpu-prices https://lantern-works-mcp.lantern-works.workers.dev/mcp
```

This adds it to the current project. Add `-s user` to use it everywhere.

### Any client

Most clients accept this shape. Some name the URL field differently: `serverUrl` in Devin Desktop, `httpUrl` in Gemini CLI’s settings file.

```json
{
  "mcpServers": {
    "lantern-works-gpu-prices": {
      "type": "http",
      "url": "https://lantern-works-mcp.lantern-works.workers.dev/mcp"
    }
  }
}
```

Zed puts servers under `context_servers` in its settings. Cline needs `"type": "streamableHttp"`, or it assumes SSE.

Client only starts local servers? Bridge it with `npx -y mcp-remote https://lantern-works-mcp.lantern-works.workers.dev/mcp`.

## Recent changes

- 2026-11-17 17:22 UTC: Tracking started: 13 tools read. Source: automatic check.

More on the History tab, with descriptions before and after for the newest entries: https://mcp.tc/i/lantern-works-gpu-prices#history (Atom feed: https://mcp.tc/i/lantern-works-gpu-prices/history.xml). It records what mcp.tc's anonymous checks saw, every 3 hours where the tool list can be read without sign-in; it is not a security review.

## Details

- Server version: 0.6.5
- MCP protocol version: 2025-11-25
- Last checked: 2026-11-18 (reachable)
- Listed: 2026-11-17
- Updated: 2026-11-17

---
Source: https://mcp.tc/i/lantern-works-gpu-prices (mcp.tc is an independent directory, not affiliated with this server's publisher). Corrections: https://mcp.tc/report
