# Langfuse MCP server

> Work with Langfuse prompts, observations, datasets, scores, evaluators and dashboards for LLM apps from your AI assistant.

- Listing: https://mcp.tc/i/langfuse
- Connect: use the server's own URL `https://cloud.langfuse.com/api/public/mcp` (with your API key); clients connect to it directly. The listing link is a page, not an MCP endpoint.
- Type: remote (Streamable HTTP)
- Auth: API key
- Category: [Observability & Incidents](https://mcp.tc/c/observability)
- Vendor: Langfuse
- Homepage: <https://langfuse.com/docs/api-and-data-platform/features/mcp-server>
- Repository: <https://github.com/langfuse/langfuse>

## About

Connects an AI assistant to a Langfuse project, the open source LLM engineering platform for tracing, prompt management and evaluation. The assistant can work with prompts, observations, datasets, scores, evaluators and dashboards for your LLM applications.

Runs as a hosted streamable HTTP endpoint on Langfuse Cloud (https://cloud.langfuse.com/api/public/mcp), with US, Japan and HIPAA region endpoints, and at your-host/api/public/mcp for self-hosted instances. Authentication uses a Basic Authorization header built from your project public and secret API keys.

## What it can do

- Read and manage prompts stored in Langfuse
- Inspect observations from LLM application traces
- Work with datasets used for evaluation
- Review and create scores
- Manage evaluators and dashboards
- Use cloud regions or a self-hosted Langfuse instance

## Example prompts

- "List the prompts in my Langfuse project and show the latest version of each."
- "Show recent observations that had low scores in Langfuse."
- "Add these test cases to my Langfuse dataset for the support bot."
- "Which evaluators are configured in my Langfuse project?"

## Install

### Claude Code

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

```bash
claude mcp add --transport http langfuse https://cloud.langfuse.com/api/public/mcp
```

2. Start Claude Code and type `/mcp`. **langfuse** should show as connected.

Langfuse needs an API key. Langfuse’s docs say which header carries it; most servers use `Authorization: Bearer <YOUR_API_KEY>`.

### Claude Desktop

1. Custom connectors can’t send this server’s key header, so use the `mcp-remote` bridge. Open **Settings → Developer → Edit Config** and add:

`claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "langfuse": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://cloud.langfuse.com/api/public/mcp"
      ]
    }
  }
}
```

2. Replace the placeholder with your key and restart Claude Desktop.

The bridge needs Node.js on your computer.

### Cursor

[Add to Cursor](<https://cursor.com/install-mcp?name=langfuse&config=eyJ1cmwiOiJodHRwczovL2Nsb3VkLmxhbmdmdXNlLmNvbS9hcGkvcHVibGljL21jcCJ9>) (opens Cursor)

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

`mcp.json`:

```json
{
  "mcpServers": {
    "langfuse": {
      "url": "https://cloud.langfuse.com/api/public/mcp"
    }
  }
}
```

Langfuse needs an API key. Langfuse’s docs say which header carries it; most servers use `Authorization: Bearer <YOUR_API_KEY>`.

### VS Code

[Install in VS Code](<https://vscode.dev/redirect/mcp/install?name=langfuse&config=%7B%22type%22%3A%22http%22%2C%22url%22%3A%22https%3A%2F%2Fcloud.langfuse.com%2Fapi%2Fpublic%2Fmcp%22%7D>) (opens VS Code)

Or from a terminal:

```bash
code --add-mcp '{"name":"langfuse","type":"http","url":"https://cloud.langfuse.com/api/public/mcp"}'
```

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

`.vscode/mcp.json`:

```json
{
  "servers": {
    "langfuse": {
      "type": "http",
      "url": "https://cloud.langfuse.com/api/public/mcp"
    }
  }
}
```

Langfuse needs an API key. Langfuse’s docs say which header carries it; most servers use `Authorization: Bearer <YOUR_API_KEY>`.

### 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": {
    "langfuse": {
      "serverUrl": "https://cloud.langfuse.com/api/public/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 langfuse --url https://cloud.langfuse.com/api/public/mcp
```

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

`config.toml`:

```toml
[mcp_servers.langfuse]
url = "https://cloud.langfuse.com/api/public/mcp"
```

Langfuse needs an API key. Langfuse’s docs say which header carries it; most servers use `Authorization: Bearer <YOUR_API_KEY>`.

### Gemini CLI

```bash
gemini mcp add --transport http langfuse https://cloud.langfuse.com/api/public/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": {
    "langfuse": {
      "type": "http",
      "url": "https://cloud.langfuse.com/api/public/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://cloud.langfuse.com/api/public/mcp`.

Langfuse needs an API key. Langfuse’s docs say which header carries it; most servers use `Authorization: Bearer <YOUR_API_KEY>`.

## Details

- Last checked: 2026-10-03 (reachable, asks for credentials)
- Listed: 2026-10-03
- Updated: 2026-10-03

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