AudD Music Recognition MCP. Identify any song from a URL or a few lyrics.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
AudD Music Recognition is a universal platform that lets your AI client identify songs from audio URLs, search lyrics by text, and fetch full music metadata.
It works by taking an audio input and returning structured data—artist, title, album, and direct streaming links for major platforms like Spotify and Apple Music.
Use it to track background music in videos or find the source of a catchy tune when you only remember a few lyrics.
What your AI agents can do
Add monitored stream
Registers an audio stream so the server can monitor it for music detection.
Get lyrics
Retrieves the complete lyrics for a song when given the track's identity.
List monitored streams
Lists all audio streams that are currently being monitored by the server.
The agent takes an audio URL and returns the track's artist, title, album, and links to streaming services.
The agent accepts a text fragment and returns the full lyrics and associated song metadata.
The agent registers an audio stream and can retrieve a list of all currently monitored streams.
The agent analyzes a long audio file, starting the recognition process at a specific time offset.
The agent takes a general text query and finds matching songs and metadata.
The agent fetches deep metadata for a known track, including release dates and record labels.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
AudD Music Recognition MCP Server: 8 Tools
Use these eight tools to monitor streams, search lyrics, and extract detailed music metadata from audio sources.
019d8418add monitored stream
Registers an audio stream so the server can monitor it for music detection.
019d8418get lyrics
Retrieves the complete lyrics for a song when given the track's identity.
019d8418list monitored streams
Lists all audio streams that are currently being monitored by the server.
019d8418recognize at time
Starts music recognition on an audio file, focusing only on a specific time offset.
019d8418recognize music
Analyzes an audio URL and returns the artist, title, album, and direct streaming links.
019d8418search lyrics
Finds full lyrics and song metadata when you provide a text fragment.
019d8418search music
Searches the entire database to find songs matching a general text query.
019d8418set stream callback url
Sets the webhook URL where the server sends results for continuous stream monitoring.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with AudD Music Recognition, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
This server lets your agent act like a music detective. It takes audio links or lyrics and gives you structured data you can use right away. You'll get the artist, the title, the album, and even direct links to Spotify and Apple Music.
recognize_music analyzes an audio URL and returns the artist, title, album, and direct streaming links.
search_lyrics finds full lyrics and song metadata when you give it a text fragment.
search_music searches the whole database for songs matching a general text query.
get_lyrics retrieves the complete lyrics for a song once you know the track's identity.
For continuous audio, add_monitored_stream registers an audio stream so the server monitors it for music detection. You can then use list_monitored_streams to see all the streams it's tracking.
If you're dealing with a long audio file, recognize_at_time starts music recognition, focusing only on a specific time offset.
If you need the results from a continuous stream, set_stream_callback_url sets the webhook URL where the server sends the results.
How AudD Music Recognition MCP Works
- 1 First, your agent calls
add_monitored_streamto tell the server which audio stream to watch. - 2 Next, the server monitors the stream. When music is detected, your agent can call
list_monitored_streamsto check the status, or userecognize_at_timefor specific time markers. - 3 Finally, the agent uses the detected data to call
get_lyricsorsearch_musicto get the full context and links.
The bottom line is, your agent handles the whole workflow—from setting up the stream monitoring to delivering the final, actionable song data.
Who Is AudD Music Recognition MCP For?
The content creator who needs background music identification for videos. The digital archivist who needs to cross-reference obscure song lyrics. The developer building a media tool that needs reliable, real-time audio fingerprinting. If your job involves figuring out 'what song is this?' for a living, this is for you.
Uses recognize_music to identify background music in video clips, ensuring proper licensing and accurate credits.
Employs recognize_at_time to pinpoint the exact start time of a song within a longer recorded mix.
Integrates search_lyrics and recognize_music into a backend workflow, providing users with instant song identification.
Uses search_music and get_lyrics to build databases of song metadata and track historical data.
What Changes When You Connect
- Find the source of any background track. Just feed the audio URL into
recognize_musicand get the artist, title, and direct streaming links. No guessing required. - Never lose a lyric again. Use
search_lyricsto find the full text for a song, even if you only remember a few lines or a chorus snippet. - Handle long files efficiently. If the song starts 3 minutes in, don't re-process the whole thing. Use
recognize_at_timeto pinpoint the track starting at a specific offset. - Manage live data feeds. Use
add_monitored_streamandset_stream_callback_urlto let your agent watch an ongoing audio stream and get notified when music changes. - Build a deep metadata layer. Use
search_musicto search by genre or artist name, and then useget_lyricsto pull the full history and associated data for that track. - Streamline development. Instead of calling five different APIs for metadata,
recognize_musicbundles the artist, title, album, and streaming links into one call.
Real-World Use Cases
The Video Editor Needs Background Music
A content producer has a video clip and needs to credit the background music. They pass the video's audio URL to recognize_music. The agent returns the track name, artist, and links. The producer then uses the metadata to ensure accurate credit placement in the video description.
The Podcast Host Needs a Quote’s Source
A podcast host hears a song snippet and only knows a key lyric. They run search_lyrics with the text fragment. The agent identifies the song and provides the full metadata, allowing the host to research the song's history for context.
The Live Streamer Needs Real-Time Tracking
A streamer needs to track music changes in a continuous feed. They call add_monitored_stream and set a callback. The agent watches the feed, and when the song changes, the callback fires, giving the streamer instant, actionable data.
The Developer Needs to Build a Feature
A developer is building a music app and needs to identify music in uploaded files. They use add_monitored_stream for the continuous feed, and recognize_music for fixed files, giving them both stream and file-based identification capabilities in one workflow.
The Tradeoffs
Searching by Artist Name Only
Trying to find a song by only submitting the artist's name to a general search field, which only returns a list of albums and requires multiple clicks to find the specific song.
→
Use search_music with a combined query (e.g., 'artist: Queen AND song: Bohemian Rhapsody') or pass an audio URL directly to recognize_music. This gives immediate, structured results without browsing a catalogue.
Ignoring Time Constraints
Running recognize_music on a 10-minute track when the song is only 30 seconds long and starts at the 8-minute mark. This wastes time and bandwidth.
→
Use recognize_at_time instead. You specify the exact time offset (e.g., 8:00) to limit the scope of the search, guaranteeing faster and more accurate results.
Assuming Data Availability
Assuming that just knowing the song title is enough to get streaming links, when in reality, the data is often incomplete or requires a secondary API call to fetch the final links.
→
Use recognize_music. This tool bundles the title, artist, and streaming links into a single response, eliminating the need for multiple follow-up calls.
When It Fits, When It Doesn't
Use this if you need to identify a track and want the full associated data (lyrics, links, metadata) in one flow. It’s perfect for content pipelines where you process audio files and need to generate reports or links immediately. Don't use it if you only need to list streams—use list_monitored_streams. And if you're building a simple search box that only takes text, you might only need search_music, but AudD gives you the stream monitoring capabilities too, which is a bigger deal.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by AudD. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 8 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Figuring out background music shouldn't require opening three different apps.
Today, if you hear a song and need to identify it, you usually open one app (like Shazam), get the name, then open a second app (like Wikipedia) to get the artist's history, and finally open a third app (like Spotify) to find the streamable link. It's a lot of copy-pasting and context switching.
With AudD, you feed the audio URL to your agent. It runs `recognize_music` and spits out the artist, title, album, and the working links—all in one go. You get the full data set, not just a name.
AudD Music Recognition MCP Server: Get the full song data instantly.
Manual workflows used to require you to run separate scripts: one to monitor the stream (`add_monitored_stream`), another to get metadata (`search_music`), and a third to fetch lyrics (`get_lyrics`). These had to run sequentially and manage complex state.
Now, the agent manages the whole state. It handles the stream lifecycle, runs the recognition, and then pulls the lyrics. It’s a single, cohesive process that just works.
Common Questions About AudD Music Recognition MCP
How do I use the `recognize_music` tool? +
You provide an audio URL. The tool returns the artist, title, album, and direct streaming links for that track. This is the main function for identifying songs.
What is the difference between `search_music` and `recognize_music`? +
recognize_music needs an audio file or URL to identify a song. search_music just takes text and searches the metadata database for matching tracks.
Can `get_lyrics` retrieve lyrics for a song I only remember a snippet of? +
No, get_lyrics requires the full song identity (title/artist) first. For snippets, use the search_lyrics tool, which is designed to find songs based on text fragments.
How does `recognize_at_time` work? +
recognize_at_time lets you start the recognition process at a specific time offset in a long audio file, which saves time and bandwidth compared to processing the whole file.
Do I need to use `add_monitored_stream` for live audio? +
Yes. You must use add_monitored_stream to register the audio source first. This tells the server to start watching the feed, allowing the agent to process changes.
How do I use the `list_monitored_streams` tool to check my current setup? +
Running list_monitored_streams shows every audio feed your AI client is currently tracking. This is useful for confirming which streams are active or debugging why a new feed isn't reporting results.
What is the purpose of the `set_stream_callback_url` tool? +
This tool establishes a webhook URL for real-time results. Instead of constantly polling, the server sends monitoring data directly to your specified endpoint when new music is detected.
Does `recognize_music` handle partial or unknown audio inputs? +
Yes, recognize_music analyzes audio content to identify the best match. It returns the artist, title, and album even if the input is muffled or only a small segment of the song.
Can I identify a song just by providing a link to an MP3 file? +
Yes! Use the recognize_music tool and provide the direct URL to the audio file. Your agent will identify the track and provide metadata and streaming links.
How do I find the lyrics for a song if I only know a few words? +
You can use the search_lyrics tool with those words. Your agent will search AudD's database for matching lyrics and return the full text and song details.
Can I get direct links to open the song on Spotify or Apple Music? +
Yes, when identifying a song or searching for one, you can request extra data that includes direct streaming links for various platforms.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
Petfinder
Largest adoptable pet database — search dogs, cats, and organizations via AI.
OverDrive Library API
Search library collections — audit books, formats, and media via AI.
Goodcall
Answer business phone calls with an AI receptionist that schedules appointments, takes messages, and never puts callers on hold.
You might also like
FishBase
Access the world's largest database of fish species. Query common names, database versions, and technical documentation for FishBase and SeaLifeBase.
Exa AI
Search and discover the web — audit semantic results and similar links via AI.
Recharge
Automate subscription commerce via Recharge — manage subscriptions, customers, and orders directly from any AI agent.