# ACRCloud Music Recognition MCP MCP > ACRCloud Music Recognition instantly identifies songs and pulls rich metadata from audio files or URLs. Use this MCP to get deep details on tracks, artists, and albums, including direct streaming links for Spotify and YouTube. ## Overview - **Category:** artificial-intelligence - **Price:** Free - **Tags:** content-recognition, music-metadata, audio-fingerprinting, streaming-links, media-analysis ## Description This connector lets your agent listen to music—whether you hand it a file link or just tell it what sounds like background noise—and figure out exactly what that track is. It pulls more than just the title; you get the artist's full discography, the album details, and even professional data points like confidence scores. If you’re building any kind of media app, this is crucial for tagging content or checking copyright metadata. You connect it via Vinkius to your preferred AI client and suddenly your agent can interact with a global music database using natural conversation. It's a huge leap from manually searching five different databases; now, all that data just flows through one place. ## Tools ### get_track_by_isrc Looks up a track's details using its unique ISRC code. ### identify_music_from_url Determines the song identity from any provided audio file URL, like MP3 or WAV. ### search_albums Finds music albums when given an artist name or album title. ### search_artists Searches and lists available artists by their professional name. ### search_tracks Finds specific songs using a mix of artist names, track titles, or partial lyrics. ## Prompt Examples **Prompt:** ``` Identify this song: https://example.com/sample.mp3 ``` **Response:** ``` Identification successful! The song is 'Blinding Lights' by The Weeknd from the album 'After Hours'. Here is the Spotify link: [Spotify Link]. ``` **Prompt:** ``` Search for tracks by 'Daft Punk'. ``` **Response:** ``` I've found several tracks by Daft Punk, including 'One More Time', 'Get Lucky', and 'Harder, Better, Faster, Stronger'. Which one would you like details for? ``` **Prompt:** ``` Get the Spotify link for the song with ISRC USAT20500447. ``` **Response:** ``` Searching... I've located the track 'Get Lucky' by Daft Punk. Here is the direct Spotify link: [Spotify Link]. It also includes metadata for Apple Music and YouTube. ``` ## Capabilities ### Identify Songs Send an audio file URL and get the song title, artist, album, and detailed metadata instantly. ### Search by Unique Code Input a track's ISRC code to pull precise and reliable metadata for that specific recording. ### Find Tracks by Name or Lyrics Query the database using only the song title, artist name, or even fragments of lyrics. ### Browse Artists and Albums Search for entire artist discographies or browse albums by title to explore related tracks. ## Use Cases ### Checking Copyright for a Video Project A content creator has background music from an unknown source. They send the audio snippet to their agent, which uses `identify_music_from_url`. The result confirms the artist and album, letting them check copyright status immediately. ### Building a Media Directory Index An archivist needs to index 1,000 unknown recordings. They feed the audio URLs into their agent, which uses `identify_music_from_url` for every file, generating a structured database of metadata and streaming links. ### Finding a Deep Cut Track A music enthusiast knows an obscure song but only remembers two lines of the chorus. They ask their agent to search using `search_tracks` with those lyrics, finding the exact track when title searches would fail. ### Cross-Verifying Track Data A developer wants to confirm a song's details. First, they use `identify_music_from_url`. Then, they take the resulting ISRC and run it through `get_track_by_isrc` for absolute metadata confirmation. ## Benefits - You don't have to manually check five different APIs. Use `identify_music_from_url` once, and you get the artist details necessary for subsequent searches using `search_albums` or `search_artists`. - The system handles deep metadata retrieval automatically. When a song is identified, it pulls not just the name but also multiple streaming links (Spotify, YouTube) in one go. - Need to verify data accuracy? Run `get_track_by_isrc` with a unique code for guaranteed metadata reliability, bypassing reliance on names or URLs. - It's powerful enough for developers. You can build structured workflows that first use `search_artists` and then pipe the results into `search_tracks`, creating complex data paths. - Forget vague searching. If you only have a few lyrics fragments, `search_tracks` lets your agent pinpoint songs when name searches fail. ## How It Works The bottom line is that your AI client takes raw audio or basic data points and outputs a complete, verifiable music profile. 1. You initiate a request, giving your agent either an audio URL or specific metadata like an ISRC code. 2. The MCP routes the data through ACRCloud's recognition service and global database lookup tools. 3. Your agent gets back structured JSON containing all the verified details: artist names, multiple streaming links, album dates, and track codes. ## Frequently Asked Questions **How does the ACRCloud Music Recognition MCP work with URLs?** It takes an audio file URL and runs advanced fingerprinting to identify the song. It tells you the artist, album, and track details without needing manual searching. **What is the difference between search_tracks and identifying from a URL?** `identify_music_from_url` works with actual audio samples to figure out the song. `search_tracks` requires you to provide keywords like names or lyrics, so it searches existing records. **Can I use get_track_by_isrc for everything?** While highly accurate, `get_track_by_isrc` only works if you already have the unique International Standard Recording Code. It's a lookup tool, not an identification tool. **Does ACRCloud Music Recognition support finding songs by lyrics?** Yes, use `search_tracks`. You can provide partial lyric fragments, and the MCP will query the database to find matching tracks. **What credentials do I need when using the `identify_music_from_url` tool?** You must provide your ACRCloud Access Key and Secret to connect this MCP. These keys are generated in your personal ACRCloud dashboard, not here. Your agent uses them to authenticate every request, ensuring only authorized calls reach the service. **Are there rate limits when calling multiple tools like `search_albums` or `search_artists`?** Yes, ACRCloud enforces usage limits for API requests. The MCP handles standard throttling and returns specific HTTP error codes if you exceed your allowed calls per minute. You'll need to check the official documentation for exact quotas. **When I run `get_track_by_isrc`, what specific metadata fields should I expect in the response?** The response includes comprehensive data points like full artist name, release date, album title, and multiple streaming links. You get structured JSON containing confidence scores and platform IDs for Spotify, YouTube, and more. **What happens when `search_tracks` or other search tools fail to find a match?** The tool doesn't throw an error; it returns a clean, empty data structure. Your AI client can then check for null results and prompt the user with helpful suggestions instead of failing entirely. **How do I get ACRCloud credentials?** Sign up at the [**ACRCloud Console**](https://console.acrcloud.com/), create a 'Music Recognition' project, and you will find your Access Key, Secret, and Host in the project settings. **Which audio formats are supported for identification?** ACRCloud supports most common audio formats including MP3, WAV, AAC, and OGG. The agent will download the file from the provided URL and send it for fingerprinting. **Can I find the ISRC code for a specific track?** Yes! When you identify a song or search for metadata, the `acr_music_search` tool returns industry-standard identifiers like ISRC, allowing you to track specific recordings across different systems.