Cyanite AI API for Indie Labels and Libraries
October 28, 2026·by TrackTag team
If you are comparing the Cyanite AI API with a way to tag an indie label or production library catalog, start with the job you need done. Cyanite combines tagging with similarity search and catalog integrations. TrackTag is a self-serve option for producing track metadata in batches, with pay-per-track credits and several ways to automate the work.
The right choice depends on whether you are building catalog search and discovery, enriching deliveries, or clearing a backlog of audio files. Here is how to compare those workflows, what pricing to clarify, and how a small team can test a practical alternative without committing to an enterprise project.
Start with the job, not the feature list
An indie label may need to enrich 80 new masters before delivery. A production library may want consistent descriptions and searchable tags across several thousand tracks. Those are related tasks, but they do not automatically require the same product.
Write down the actual output you need:
- Catalog discovery: Can staff find tracks by reference audio, mood, or a written brief?
- Track enrichment: Do you need genres, moods, instruments, vocals, themes, tempo, key, or descriptions?
- Delivery workflow: Can results be exported in the format your team or client needs?
- Automation: Should analysis start when a file arrives, or do you need metadata returned inside your own platform?
- Budget: Is a monthly platform or API commitment reasonable for the number of tracks you process?
If finding similar music inside a large catalog is the central requirement, Cyanite deserves a serious evaluation. If the main requirement is to analyze audio files and move the results into existing workflows, compare the costs and setup effort for that specific job.
What the Cyanite AI API and similarity search are built to do
Cyanite offers audio analysis, auto-tagging, descriptions, similarity search, free text search, and integrations with catalog platforms. Its API supports integrating those capabilities into a product, while its web app offers a way to work without building a custom interface. Cyanite’s FAQ also lists integrations with platforms including Synchtank, Harvest Media, Reprtoir, and DISCO workflows. (cyanite.ai)
The Cyanite AI similarity search is especially relevant to music libraries. It is designed to find tracks that resemble an audio reference, and the API documentation describes filtering results by characteristics such as BPM, key, voice, and genre. Search is valuable when a supervisor or curator needs to discover candidates within a catalog, not just describe each track. (api-docs.cyanite.ai)
Cyanite also announced Tagging 2.0 in August 2026, adding or expanding areas such as lyric analysis, vocal style, use case, and structural segmentation. Its announcement describes an API intended to support catalogs ranging from thousands to millions of tracks. That is meaningful context for a team evaluating search and tagging at enterprise scale. (cyanite.ai)
The tradeoff for a smaller operation is usually not whether Cyanite has useful capabilities. It is whether you need the complete discovery and integration setup, and what that setup costs for your volume. Cyanite’s current FAQ says larger catalog pricing is tailored and may include an API subscription, usage-based analysis fees, and search subscriptions. A separate Cyanite API article lists €290 per month as an API usage fee. Treat that figure as a published reference, not a universal price for every plan, and request a current quote for your requirements. (cyanite.ai)
When TrackTag fits a label or library better
TrackTag is designed for teams that want to start analyzing tracks themselves, see the credit cost upfront, and export the results without first planning a custom enterprise rollout. In TrackTag Studio, the batch audio analyzer, you can drop in audio files and get up to 35 fields per track, including BPM, key, genres, subgenres, moods, emotions, themes, occasions, instruments, vocals, song structure, production and mixing notes, and a full description.
You can choose between two levels:
- Core uses one credit per track and returns nine fields intended to help identify and find a track, with keyword tags included.
- Ultra uses two credits per track and returns all 35 fields, including the written description.
Both levels use the same engine on the same audio, so they are equally accurate. Core gives you fewer fields, not a lower quality analysis. If a track later needs the full description or other Ultra fields, reanalyze it for the one credit difference.
That makes it possible to set a deliberate policy. For example, an indie label might run Core across every new delivery, then use Ultra for tracks pitched for sync or tracks that need richer catalog descriptions. A library preparing a large back catalog could start with Core and upgrade only the records where more detail is useful.
TrackTag’s published credit packs are $20 for 50 credits, $89 for 250, $149 for 1,000, and $499 for 5,000. There are free credits to start, and the packs are one-time purchases rather than subscriptions. At one credit per Core track, the 1,000-credit pack covers 1,000 Core analyses; Ultra uses two credits per track. There is also a $49 per month unlimited plan that uses your own Google AI key. Choose based on how much you analyze and whether recurring or one-time spend suits your workflow.
For export, choose TXT, CSV or Excel, XML, JSON, schema.org JSON-LD, Markdown, or PDF. You can export the whole batch, a hand-picked subset, or separate files for each track in a ZIP. That is useful when different artists, clients, or catalog projects need separate handoffs.
A workable catalog workflow, from backlog to delivery
Suppose your label has 600 released masters in folders on a local drive, plus a steady stream of new deliveries. A simple operating process could look like this:
1. Connect the source folder. My Library indexes local files and shows which are Tagged or Untagged. The initial scan reads file names and sizes on your machine. It does not upload audio unless you explicitly analyze a track.
2. Set a field policy. Use Core when searchable keywords and essential track fields are enough. Use Ultra when staff need the full set of fields and written descriptions.
3. Work through the untagged backlog. Sort the library view, use folder aliases to distinguish projects or clients, and select “Analyze untagged” to process the outstanding files.
4. Export for the next system. Use CSV or Excel for review, JSON for a software workflow, or separate per-track files when handing results to different people.
5. Automate new arrivals. Use TrackTag’s Zapier app to connect an “Analysis Finished” trigger, an “Analyze Track” action, or a “Get Analysis” lookup with tools such as Google Sheets, Dropbox, Airtable, Notion, or Slack. The TrackTag Zapier setup guide walks through the connection.
Folder access in the browser works with Chrome, Edge, and Brave. Safari and Firefox do not support that folder access, so the desktop app for Mac and Windows is an option when you need persistent connected folders or long batches running in their own window.
For measured audio fields, TrackTag says BPM is derived from the signal using Beat This!, an open-source beat-tracking model it runs on its own servers. The model reads the exact BPM on 82% of the 999 songs in the GTZAN benchmark; if a track has no steady beat, it returns “No BPM” instead of inventing a number. Key is measured by TrackTag’s signal-analysis engine. If these fields matter for matching or delivery, read how BPM and key detection work.
API, MCP, and Zapier: choose the right automation route
Cyanite has an API and integrations with catalog platforms, which can be a strong fit when you want its analysis and search embedded in a larger system. TrackTag’s public API is a direct option for sending a track and receiving the same analysis as JSON. It uses the same credit balance as Studio, with defaults of 10 requests per minute and 2,000 analyses per day per key. Keys are created in Studio. See the TrackTag developer documentation for API details.
The API can suit a marketplace that analyzes uploads, a label enriching incoming deliveries, or a software team that wants analysis in its own interface. If your immediate need is to connect existing apps without code, the Zapier app may be a better starting point than a custom integration.
TrackTag also provides an MCP server that runs on your machine. Compatible assistants such as Claude Desktop, Cursor, and Claude Code can analyze a track, search local audio files by name, or check the credit balance from within a conversation. We could not verify a public Cyanite MCP server or a Cyanite listing in Zapier’s app directory in the sources reviewed for this article, as of October 2026. Check directly with Cyanite before making that a deciding factor.
The key distinction is practical: Cyanite combines analysis with catalog search and offers enterprise scale, while TrackTag offers a self-serve path with batch processing, explicit credit packs, a public API, MCP, and Zapier. For a direct feature and workflow comparison, see TrackTag vs Cyanite.
A simple way to decide
Choose Cyanite when similarity search, free text search, a supported catalog integration, or processing at very large scale is central to the project. Ask for the full quote, including analysis, search, integration, and any subscription components. Test the catalog you actually need to search, since usefulness depends on the contents and size of that catalog.
Choose TrackTag when you mainly need to tag and describe files in batches, make consistent exports, process a local folder, or connect analysis to a lightweight automation. Start with a small representative batch: include instrumental and vocal tracks, several genres, alternate production styles, and tracks with and without a steady beat. Compare the resulting fields against your real delivery or search needs before processing the full catalog.
For a production library, judge the result by whether a music supervisor can retrieve a useful track from the tags and description, not by how many fields appear on a feature page. For a label, judge it by whether staff can process new deliveries and hand metadata to the next system without adding a costly monthly commitment they do not use.
FAQ
What does Cyanite use AI for?
Cyanite uses its music analysis and search tools to generate tags and descriptions, find similar tracks, and support free text search across music catalogs. Its announced Tagging 2.0 capabilities include lyric analysis, vocal style, use case, and structural segmentation. (cyanite.ai)
Does Cyanite have an API for music tagging?
Yes. Cyanite documents an API for its music tagging product, and its product information also describes API access for analysis and search. Business pricing is customized, so confirm the current charges for your catalog and selected features before budgeting. (docs.cyanite.ai)
Is Cyanite’s similarity search useful for a small catalog?
It can be, but similarity search returns matches from the catalog it can search, so the best available result may still be a weak match. Cyanite says similarity scores help users assess that difference and suggests a catalog of at least 1,000 songs for its similarity search workflow. (cyanite.ai)
Can I analyze a local music folder with TrackTag?
Yes. TrackTag My Library indexes a connected local folder and marks files as Tagged or Untagged. The scan reads file names and sizes locally; a track is uploaded only when you explicitly choose to analyze it.
Tag your whole catalog with AI
BPM, key, genre, moods, instruments and keywords: 30+ fields per track, exported ready for libraries.
Open TrackTag Studio →