# Xeno-canto MCP > Search recordings with the search_recordings tool. Connect directly to the world’s largest open database of bird sounds, giving your agent access to over 800,000 global wildlife audio recordings. Filter results by species, country, genus, or quality grade for bioacoustics research and identification. ## Overview - **Category:** audio-music - **Price:** Free - **Tags:** ornithology, bioacoustics, wildlife, bird-songs, nature ## Description This MCP gives your AI client direct access to the Xeno-canto database, a massive collection of bird sounds from around the globe. Instead of manually navigating complex websites, your agent can query this data using natural language or advanced filters. You can search across every continent for specific species, retrieve metadata like recording date and location, or narrow down results by quality grade. This is crucial for ornithologists and researchers who need to analyze large bioacoustic datasets quickly. The power comes from combining searches; you might run one query for the genus *Turdus* in France, then another query for a specific call type, all through your agent. Because every tool call runs inside Vinkius's secure environment, you get full visibility into what data flowed and how many tokens were used by checking the AI Analytics dashboard—nothing happens in the dark. ## Tools ### search_recordings Searches the Xeno-canto database to find specific bird sound recordings using filters like genus, species, or location. ## Prompt Examples **Prompt:** ``` Search for recordings of the Common Blackbird in the Netherlands. ``` **Response:** ``` I've found several recordings for the Common Blackbird (Turdus merula) in the Netherlands. The top results include high-quality songs recorded in Gelderland and Utrecht. Would you like to see the details for a specific recording? ``` **Prompt:** ``` Find high-quality recordings of the genus Turdus in France using quality grade A. ``` **Response:** ``` Searching for high-quality (grade A) recordings of Turdus in France... I've retrieved 15 recordings, including Turdus philomelos and Turdus merula. All these recordings are rated 'A' for excellent sound clarity. ``` **Prompt:** ``` Get the second page of results for 'owl' recordings. ``` **Response:** ``` Fetching page 2 of the search results for 'owl'... I have loaded the next set of recordings, featuring various species like the Barn Owl and Tawny Owl from different global locations. ``` ## Capabilities ### Identify specific species calls Search for recordings using filters like genus or species name to pinpoint exact bird vocalizations. ### Filter by geography and time Restrict searches based on the country, location, or date of recording to narrow down global datasets. ### Analyze data quality Query results specifically for certain sound types (song vs. call) or filter by recorded audio clarity/grade. ### Handle large result sets Navigate through thousands of records using built-in pagination support without losing track of your search progress. ## Use Cases ### Comparing subspecies calls An ornithologist needs to compare the vocalizations of two closely related bird species across three different continents. They use their agent to repeatedly run search_recordings, filtering by both species name and country/continent, building a comparative dataset rapidly. ### Verifying historical recordings A researcher is checking data from 1950s expeditions. They use the MCP to query search_recordings, applying date filters and specific location metadata to ensure only historically accurate files are retrieved for a paper. ### Identifying unknown calls A nature enthusiast records an unknown bird call in their backyard. They feed the agent details about the potential region and use search_recordings, refining filters by genus until they find a match with high certainty. ## Benefits - Pinpoint species vocalizations: Use search_recordings to filter recordings by specific genus, ensuring you only get the data relevant to your research. - Global coverage in one query: Access sound data from every continent and thousands of subspecies without geographic limitations. This is key for comparative studies. - Deep metadata retrieval: Get more than just audio files; retrieve critical details like location, date, recordist, and specific sound type (call vs. song) immediately. - Efficient large-scale analysis: The built-in pagination support means you can process massive result sets page by page without running into data limits. - Targeted research filtering: Limit your search to only 'A' quality grade recordings or specific sound types, cutting through noise and focusing on usable audio. ## How It Works The bottom line is you get structured access to millions of audio data points without touching a web browser or API key. 1. Ask your agent to perform a bioacoustic query, specifying parameters like species name and country. 2. The MCP executes the request, querying the massive Xeno-canto dataset against those filters. 3. You receive paginated results containing recordings, metadata (location, date), and classification details. ## Frequently Asked Questions **How can I find recordings of a specific bird species?** Use the `search_recordings` tool with the species name or scientific name (e.g., 'Common Blackbird' or 'Turdus merula'). You can combine this with other filters like country or quality. **Can I filter results by recording quality or location?** Yes! The `search_recordings` tool supports Xeno-canto syntax. Use `cnt:france` for location or `q:A` for the highest quality recordings within your query string. **Is there a limit to the number of results I can retrieve?** The `search_recordings` tool returns results in pages. You can use the `page` parameter to navigate through the database, with each page containing up to 500 recordings. **What is the setup process for using the `search_recordings` tool?** You don't need to worry about credentials. Because this MCP accesses the public Xeno-canto database, no API key or special authentication is needed to run `search_recordings`. Just connect your agent and start querying. **Beyond the audio file, what metadata does `search_recordings` retrieve?** The tool returns detailed metadata for every match. You get location data, the recording date, who recorded it (the recordist), and even whether the sound is classified as a song or a call. **How can I use advanced query syntax with `search_recordings`?** You can absolutely use complex filters. The tool accepts queries specifying criteria like genus, country, and quality grade (e.g., 'gen:Turdus cnt:france q:A'). This lets you narrow down results dramatically. **If my query using `search_recordings` returns nothing, what should I do?** It won't throw an error; it will simply return an empty set. If that happens, try widening your search parameters or simplifying the filters you passed to the tool. **Does `search_recordings` handle massive datasets and pagination?** Yes, it's built for large scale data. The MCP includes native page management support, letting you navigate through huge result sets without hitting artificial limits.