Bocha Web SearchLABSCO SUMMARY
Search results built for a model to read, with live values pages cannot give.
The summaries come back long enough to reason over rather than as a one-line snippet, which is the difference between search a model can answer from and search it has to click through. The cards cover the questions where the answer changes by the minute — a share price or today's weather is rarely reliable on whatever page a query surfaces.
What it is
A search server from Bocha with two tools. Web search returns page titles, links, summaries, publication times and site names from a very large index. AI search adds structured cards on top, selected from what the query means — weather, encyclopaedia, calendar, trains, stocks and dozens more.
What you get
- Web results with title, link, summary, publication time and site name
- A time window on the search — a specific date, a date range, or the last day, week, month or year
- Control over how many results come back
- Structured cards alongside the results for domains where a page cannot carry the current value: weather, stocks, exchange rates, fuel prices, trains, almanac, medical and encyclopaedia among others
- Coverage that reaches beyond web pages into partner content — video, news, encyclopaedia, travel, hospitality and academic sources
Requirements
An API key from the Bocha open platform, supplied as an environment setting. The server runs as a Python package launched with uv from a clone of the repository.
Setup effort
One command plus a key — uvx bocha-search-mcp, then supply credentials
USE CASES
01
Answering a question about the Chinese-language web
Search across the index and get back titles, links and summaries with enough text to answer from directly.
02
Scoping a search to what is recent
Restrict the window to the last day, week or month, or to a date range you name, when only current pages count.
03
Getting a value a web page cannot give
Use the search that returns structured cards, so a stock price, an exchange rate or today's weather comes back as data rather than as a stale page.
04
Feeding a model enough to work with
Raise the result count and let the summaries, rather than a stack of links, carry the answer.
2 tools
Listed in project docs. Your client may show a different list.
Bocha AI SearchBEFORE YOU INSTALL
What to know first
Tools you get
2 tools named in the project's documentation.
Requires
- uvx on PATH
- BOCHA_API_KEY in your environment
INSTALLATION
Connect your client
Server setup
TERMINAL
uvx bocha-search-mcpMCP Clients
Claude Desktop
Download the code
The Claude Desktop entry the project publishes runs the server from a local copy with uv (Python 3.12 or later).
CLAUDE DESKTOP · TERMINAL
git clone https://github.com/BochaAI/bocha-search-mcp.gitAdd to claude_desktop_config.json
<YOUR_PATH_TO_BOCHA_SEARCH_MCP> is the absolute path of the folder you just cloned. <YOUR_BOCHA_API_KEY> is the API key you get after signing in to the Bocha open platform (open.bochaai.com).
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"bocha-search-mcp": {
"command": "uv",
"args": [
"--directory",
"<YOUR_PATH_TO_BOCHA_SEARCH_MCP>",
"run",
"bocha-search-mcp"
],
"env": {
"BOCHA_API_KEY": "<YOUR_BOCHA_API_KEY>"
}
}
}
}
