Google Surf MCP
Google Surf MCP is an MCP server that adds web browsing tools to AI assistants such as Claude Desktop, Claude Code and Cursor. Web, academic and code search with graph RAG, data lineage, ontology and cross-project schema links. It has 301 GitHub stars, is released under the MIT license and runs locally with npx -y google-surf-mcp.
github.com/HarimxChoi/google-surf-mcp (opens in a new tab)
Data as of from the GitHub API and the official MCP registry; description from the project's repository. Report an error
- Needs API key
- Web & browsing
- MIT
- Active
Install Google Surf MCP
Generated from the server's MCP registry entry. Replace your-value with your own values.
Claude Desktop
{
"mcpServers": {
"google-surf-mcp": {
"command": "npx",
"args": [
"-y",
"google-surf-mcp"
],
"env": {
"SEARCH_API": "your-value",
"GITHUB_TOKEN": "your-value"
}
}
}
}Settings > Developer > Edit Config. macOS: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\. Restart Claude Desktop afterwards.
Claude Code
claude mcp add --env SEARCH_API=your-value --env GITHUB_TOKEN=your-value --transport stdio google-surf-mcp -- npx -y google-surf-mcpCursor
{
"mcpServers": {
"google-surf-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"google-surf-mcp"
],
"env": {
"SEARCH_API": "your-value",
"GITHUB_TOKEN": "your-value"
}
}
}
}Project file; use ~/.cursor/mcp.json to enable it in every project.
VS Code
{
"inputs": [
{
"type": "promptString",
"id": "search_api",
"description": "SEARCH_API",
"password": true
},
{
"type": "promptString",
"id": "github_token",
"description": "GITHUB_TOKEN",
"password": true
}
],
"servers": {
"google-surf-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"google-surf-mcp"
],
"env": {
"SEARCH_API": "${input:search_api}",
"GITHUB_TOKEN": "${input:github_token}"
}
}
}
}Config formats checked against the official docs on Oct 7, 2026: modelcontextprotocol.io (opens in a new tab), code.claude.com (opens in a new tab), cursor.com (opens in a new tab), code.visualstudio.com (opens in a new tab).
Environment variables
Variables the server reads at startup.
| Name | Required | Description |
|---|---|---|
SEARCH_APIsecret | No | SearchApi API key. Required only for SearchApi primary or fallback modes. |
SURF_SEARCH_PROVIDER | No | Provider for search and search_parallel: browser, searchapi, or fallback. Defaults to browser. |
SURF_SCHOLAR_PROVIDER | No | Provider for scholar_search: browser, searchapi, or fallback. Defaults to browser. |
SURF_BROWSER_ENGINE | No | Browser engine: auto, native, or playwright. Auto uses one hidden native Chrome broker shared across local MCP sessions. |
CHROME_PATH | No | Absolute path to the Chrome binary. Auto-detected on Windows/macOS/Linux when omitted. |
SURF_PROFILE_ROOT | No | Directory for the warm Chrome profile. Defaults to ~/.google-surf-mcp. |
SURF_RESEARCH | No | Enable Google Surf Auto Ontology: local project memory, evidence capture, and hybrid retrieval. Defaults to true. The SURF_RESEARCH setting name is unchanged. |
SURF_RETRIEVAL_MODE | No | Shared route for search and search_parallel: live or hybrid. Defaults to hybrid. |
SURF_RESEARCH_ROOT | No | Embedded SurrealDB data directory. Defaults to /research. |
SURF_RESEARCH_VECTOR_MODEL | No | Local model for HNSW retrieval and shared reranking. Defaults to Xenova/multilingual-e5-small; use off to disable. |
SURF_RESEARCH_VECTOR_LOW_MEMORY | No | Disable ONNX CPU memory arena and memory pattern. Defaults to true. |
SURF_RESEARCH_VECTOR_THREADS | No | ONNX intra-op thread count from 1 to 16. Defaults to 4. |
SURF_RESEARCH_REPO_AUTO | No | Sparse-index at most one eligible small GitHub repository per search call. Defaults to true. |
SURF_RESEARCH_REPO_AUTO_MAX_MB | No | Maximum searchable repository source-text size in MB. Assets are excluded. Defaults to 20. |
SURF_RESEARCH_REPO_AUTO_MAX_FILES | No | Maximum repository source file count for automatic indexing. Defaults to 2000. |
SURF_RESEARCH_BROKER_IDLE_MS | No | Shared Auto Ontology broker idle lifetime in milliseconds. Defaults to 60000. |
SURF_RESEARCH_READ_CONCURRENCY | No | Maximum concurrent Auto Ontology broker reads. Defaults to 4. |
SURF_RESEARCH_QUERY_TIMEOUT_MS | No | Timeout per embedded SurrealDB query in milliseconds. Defaults to 120000; timed-out retrieval lanes are reported as partial. |
GITHUB_TOKENsecret | No | Optional token that raises GitHub API limits for repository inspection. |
SURF_LOCALE | No | Browser locale, e.g. en-US. |
Tools (5)
Parsed from the Tools section of the README; check the repository for the current list.
scholar_searchFind paper-specific Google Scholar metadata.
extractRead a web page, PDF, or repository from an already-known URL.
project_memory_searchLocal-only search of indexed project knowledge; no live web search.
project_memoryManage projects, save research records, index user-approved local folders, and export graphs.
healthInspect search and local runtime status.
About Google Surf MCP
Google Surf MCP searches the web, papers, and codebases, then uses Auto Ontology to organize what you read into a local knowledge graph. Start without an API key or a separate database server, and build a personal knowledge base from connected sources and project records.
- google-search
- mcp
- model-context-protocol
- playwright
- academic-research
- fetch
- pdf-processing
- self-healing
Similar MCP servers
More web & browsing MCP servers
Autocad MCP
U-C4N/Autocad-MCP
AutoCAD MCP server for AI agents (Claude, Cursor, any MCP client): 240+ tools to draw, edit and read DWG/DXF drawings.
Active · last commit Nov 6, 2026
Web & browsingPythonBrowsentic
imshaikot/browsentic
Agentic browser harness: your agent CLI in your real, signed-in browser, from a side panel or MCP.
Active · last commit Nov 6, 2026
Web & browsingTypeScriptsquirrelscan
squirrelscan/squirrelscan
Website QA for your coding agent: audit SEO, performance, security, accessibility over MCP.
Active · last commit Nov 16, 2026
OfficialWeb & browsingTypeScriptPagesift
YOLOW92/pagesift
Self-hosted web scraper, crawler and document parser for AI agents: URLs, PDFs with OCR and Office files to clean Markdown/JSON.
Too new to judge · created Oct 20, 2026
Web & browsingRustSpotifyscraper
AliAkhtari78/SpotifyScraper
Public Spotify metadata, lyrics and podcasts for LLM agents.
Maintained · last commit Jun 26, 2026
Web & browsingPythonClearcote Browser
clearcotelabs/clearcote-browser
Drive the Clearcote Chromium browser from any MCP client: read pages, click, fill, screenshot.
Active · last commit Nov 13, 2026
Web & browsingPython
How do I install the Google Surf MCP MCP server?
Add the command npx -y google-surf-mcp to your MCP client: put it in claude_desktop_config.json for Claude Desktop, run claude mcp add for Claude Code, or add it to .cursor/mcp.json (Cursor) or .vscode/mcp.json (VS Code). The snippets on this page are ready to paste.
Is Google Surf MCP free?
The server is open source under the MIT license, so running it is free. It needs credentials (SEARCH_API and GITHUB_TOKEN) for the service it connects to, which may require a paid account.
Is Google Surf MCP actively maintained?
The most recent commit was on Nov 16, 2026. The latest release is v1.1.6, published Nov 7, 2026. appsgit only lists MCP servers with a commit in the last six months.