FreqBlog vs Musicae
When Spotify deprecated /audio-features, /recommendations and /related-artists in November 2024, it left a straightforward buying question: which replacement do you wire in? The honest answer depends on three things — whether every track you touch already carries a Spotify ID, how much you care where the numbers come from, and who stands behind the service if something breaks. This page compares FreqBlog and Musicae on facts you can verify yourself today.
At a glance
Every cell below is a currently-verifiable public fact, with Musicae's cells drawn from its own site (api.musicae.io), Medium post and Apify profile, fetched July 2026. Catalogue size and headline price are deliberately not compared here — both services are in a similar range and neither wins the buyer anything.
| Dimension | FreqBlog | Musicae |
|---|---|---|
| Track identification | Look up a track's audio features directly by name + artist — also ISRC, MBID or Spotify ID. No prior ID lookup required. | Audio features and DJ analysis are “resolve[d] by Spotify ID or ISRC.” |
| How values are produced | Independently computed with Essentia from the source audio. | Site states analysis is “precomputed from Spotify audio-feature data.” |
| Hosted MCP server (for LLMs / agents) | Yes — hosted endpoint at mcp.freqblog.com/mcp. |
No hosted MCP server listed on its site. |
| Uncovered track | On-miss backfill: request a track that isn't in the catalogue and it is ingested on demand. | Serves from a precomputed, cached catalogue; no on-miss backfill described. |
| Operator disclosure | Operator publicly named: Steven Birring, ICO-registered UK sole trader (trading as Stackbase / FreqBlog), Nuneaton, UK. | No company, operator or founder named on its site. |
| Access & billing | Direct API with its own free tier (1,000 requests/month, no card) and dashboard. | “Billing is handled by the marketplaces” — sold via RapidAPI and Apify. |
When Musicae is the better fit
There are real cases where Musicae is the more natural choice, and it's worth being straight about them:
- You already have Spotify IDs everywhere and want a drop-in swap. Musicae mirrors Spotify's audio-features schema so you migrate with a single base-URL change — if your pipeline is already ID-keyed, that's the least-friction path.
- You want no-code bulk runs. Its Apify actors let you paste IDs/ISRCs, export JSON/CSV/Excel and schedule recurring runs without writing any integration code.
- You want DJ-software library exports out of the box. Musicae advertises Rekordbox, M3U and cuesheet exports plus popularity trajectory/momentum signals — handy if those specific outputs are your goal.
Why teams pick FreqBlog
- No Spotify ID required. Resolve a track by name + artist (or ISRC, MBID, Spotify ID). You don't have to run an ID-lookup step first, which matters when your data starts from tags, filenames or spreadsheets rather than Spotify.
- Values you can attribute. BPM, key (name + Camelot + Open Key), energy, mood, genre and the rest of the 42 fields are computed with Essentia from the source audio — not re-served Spotify values.
- Built for agents. A hosted MCP server (
mcp.freqblog.com/mcp) lets an LLM or agent call audio-feature and set-building tools directly. - Gaps close themselves. On-miss backfill means a track you request but that isn't yet covered gets ingested, rather than simply returning empty.
- A named, accountable operator. Run by Steven Birring, an ICO-registered UK sole trader (Nuneaton) — a real person you can identify and reach, with public ICO registration.
- A gentle on-ramp. Free tier of 1,000 requests/month with no card, and low per-request pricing (~£0.17 per 1,000) billed directly.
FAQ
Do I need a Spotify ID to use FreqBlog?
No. You can resolve a track by name + artist, or by ISRC, MusicBrainz ID (MBID) or Spotify ID. Musicae's audio-features and analysis endpoints, by contrast, resolve by Spotify ID or ISRC.
Where do FreqBlog's numbers come from?
They are computed independently with the open-source Essentia toolkit from the source audio. Musicae's own site describes its analysis as “precomputed from Spotify audio-feature data.”
Can I call it from an LLM or agent?
Yes — FreqBlog runs a hosted MCP server at mcp.freqblog.com/mcp, so tools like audio-feature lookup and harmonic set-building are available to agents without a custom wrapper. Musicae does not list a hosted MCP server on its site.
What happens if a track isn't in the catalogue yet?
FreqBlog's on-miss backfill ingests the requested track on demand and returns its features once processed, rather than returning nothing.