Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@bocha-mcp search for latest AI news this week"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP server for Bocha Search API, providing web search and AI search capabilities.
Web Search: Search the web using Bocha Web Search API (default)
AI Search: Search using Bocha AI Search API with multimodal cards and AI-generated answers (only when specifically requested)
Related MCP server: LangSearch MCP Server
Clone this repository
Install dependencies:
npm installCopy .mcp.example.json to .mcp.json and add your Bocha API key:
cp .mcp.example.json .mcp.jsonEdit .mcp.json and replace YOUR_BOCHA_API_KEY_HERE with your actual API key
Add the following to your Claude configuration:
{
"mcpServers": {
"bocha-search": {
"command": "node",
"args": ["/path/to/bocha-mcp/index.js"],
"env": {
"BOCHA_API_KEY": "your_api_key_here"
}
}
}
}web-search: Search the web using Bocha Web Search API
query: Search query (supports natural language)
count: Number of results (1-50, default 10)
freshness: Time range filter (noLimit, day, week, month, year)
summary: Whether to include text summary (default true)
ai-search: Search using Bocha AI Search API
query: Search query
count: Number of results (1-50, default 10)
freshness: Time range filter (noLimit, day, week, month, year)
answer: Whether to generate AI answer (default false)
stream: Whether to use streaming response (default false)
BOCHA_API_KEY: Your Bocha API key (required)
ISC
Search using Bocha AI Search API. Returns web results, images, multimodal cards (weather, calendar,百科, etc.), AI-generated answers, and follow-up questions. Supports streaming. Use only when specifically requested by the user.
| Name | Required | Description | Default |
|---|---|---|---|
| count | No | Number of results to return (1-50, default 10) | |
| query | Yes | Search query | |
| answer | No | Whether to generate AI answer | |
| stream | No | Whether to use streaming response | |
| freshness | No | Time range filter: noLimit, day, week, month, year | noLimit |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden of disclosing behavior. It reveals the tool returns diverse result types and supports streaming, which goes beyond a generic search description. It does not mention limitations or error conditions, but for a read-only search tool the core behavior is well-covered.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact and front-loaded: it starts with the core action, then enumerates outputs, mentions streaming, and ends with a usage constraint. Each sentence contributes distinct, useful information with no repetition or filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has 5 parameters all documented in the schema and no output schema, the description sufficiently explains the return types and adds usage context. It could have mentioned the freshness filter or count behavior, but those are already in the schema, so the description is adequate for a search tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% with parameter descriptions, so baseline is 3. The description adds meaning by explaining that 'AI-generated answers' and 'follow-up questions' are part of the output, which elaborates the 'answer' parameter's effect. It also mentions streaming, reinforcing the 'stream' parameter usage.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool performs a search via Bocha AI Search API, a specific resource distinct from the sibling 'web-search'. It also lists concrete output types (web results, images, multimodal cards, AI answers, follow-up questions), making the purpose unambiguous.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides an explicit usage rule: 'Use only when specifically requested by the user.' This is a clear contextual guideline, though it does not directly name alternatives or contrast with the sibling tool. The rule effectively narrows the appropriate use case.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Search the web using Bocha Web Search API. Returns web pages, images, videos with titles, URLs, snippets, and summaries. Supports natural language queries, time range filtering, and site-specific search.
| Name | Required | Description | Default |
|---|---|---|---|
| count | No | Number of results to return (1-50, default 10) | |
| query | Yes | Search query (supports natural language) | |
| summary | No | Whether to include text summary in results | |
| freshness | No | Time range filter: noLimit, day, week, month, year | noLimit |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden of disclosing behavior. It explains what is returned and supported filters, but it does not disclose rate limits, potential side effects, or limitations. This is adequate but not outstanding.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and front-loaded, consisting of two sentences that clearly state the purpose and key capabilities without unnecessary detail. Every word earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given that there is no output schema and no annotations, the description must explain return values and use context. It does so reasonably by listing returned content types and filters, but it could be improved by mentioning pagination or edge cases. Overall, it is mostly complete for a web search tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the baseline is 3. The description adds value by explaining natural language queries, time range filtering, and site-specific search, which clarifies how the parameters are intended to be used beyond the schema's individual descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly identifies the tool as a web search via Bocha API, specifying the types of results returned. However, it does not distinguish this tool from the sibling 'ai-search' tool, so it does not achieve the highest score.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
There is no guidance on when to use this tool versus the sibling 'ai-search' tool, nor any exclusions or alternative suggestions. The description only lists features and leaves the agent to infer appropriate usage.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v1.0.0ai-searchweb-searchScored across 2 tools
Both tools are search-based and can return similar result types, creating potential ambiguity. The AI search tool explicitly notes it should only be used when requested, which helps, but the boundary between general web search and AI-enhanced search is not always obvious.
Both tool names follow a consistent pattern: a type prefix followed by 'search' (web-search, ai-search). This makes the naming predictable and easy to understand.
With only two tools, the server is on the lean side, but for a focused search API server, this is reasonable. Each tool serves a distinct search mode, so the count feels appropriate for the scope.
The tool surface covers general web search and AI-powered search, including images, videos, and multimodal cards. No obvious missing functionality for the stated purpose of a search API.
MCP server for Google search results via SERP API
Serper MCP — wraps the Serper Google Search API (serper.dev)
Flight search MCP server providing search, pagination, and itinerary details for AI assistants.
