Spotify closed its audio-features endpoint to new apps in 2024. We compared every replacement — MeloData, GetSongBPM, Cyanite, Soundcharts — on features, price, and developer experience.
The number every comparison article buries. Verified against published pricing pages 2026-09-06; MeloData re-checked 2026-10-06. Equivalent commercial volume tier — not free trials, not enterprise quotes.
| API | Cheapest commercial tier | Per 1k requests | vs FreqBlog Pro |
|---|---|---|---|
| FreqBlog — Pro | £129/mo for 750k | £0.17 | — |
| MeloData — Scale, annual | $2,512/yr for 1M/mo (~$209/mo) | $0.21 | ~1.15× more vs Pro on annual billing — close |
| MeloData — Scale, monthly | $299/mo for 1M | $0.30 | ~1.4× more (monthly vs monthly) |
| AudD — volume tier | $1,800/mo for 500k | $3.60 | ~17× more |
| AudD — entry tier | $5/1k pay-as-you-go | $5.00 | ~23× more |
| MusicAPI.com — metered | €0.60/1k, €500/mo minimum | €0.60 | ~3× more (and a €500/mo floor) |
| Cyanite — published calculator | €290/mo access + per-track usage | From €0.30/track tagging | Priced per track, not per request — and upload-only |
| Soundcharts — API | From $50/mo for 10k queries | $5.00 | Different category (industry intel, not audio features) |
| GetSongBPM | Free with mandatory backlink | $0 | Cheaper, but BPM+key only — no mood/energy/genre/ISRC |
Spotify /audio-features |
Closed to new apps, Nov 2024 | — | — |
USD→GBP at $1.25/£ (mid-2026). FreqBlog has lower tiers too (Hobbyist £9.99/mo = £0.67/k, Starter £39/mo = £0.26/k); above is the cheapest-vs-cheapest commercial comparison. GetSongBPM is genuinely cheaper but only returns BPM and key — not a feature-equivalent comparison. MeloData is the closest on price: its Scale plan is within ~15% of FreqBlog Pro per 1,000 when both are billed annually (FreqBlog annual saves 15%), so choose between those two on what a call returns and how you look tracks up, not on price.
/v1/audio-features/{id} — host-swap, identical path + JSON structure (numeric values use open-source analysis; see threshold guide). Raw Spotify track IDs resolve per request by matching the track's title (we don't store Spotify IDs) — a title several artists share is ambiguous and misses rather than guessing; for reliable coverage, look up by ISRC (/v1/audio-features/{ISRC} — your Spotify track object already carries it at external_ids.isrc) or by name (/lookup?track=&artist=) instead.The closest replacement for Spotify's audio-features — not just for English-language pop. No ISRC, no Spotify token, no per-track pricing — queryable by name, MCP-native, and actively maintained.
A solid, focused product. The main difference is shape: you resolve a recording first, then fetch its features by ISRC, where FreqBlog returns the features from a single name lookup.
Great for simple BPM/key lookups but missing everything else. If all you need is tempo detection, it's a decent free option.
Purpose-built for mood and genre tagging at high accuracy. Excellent if that's your exact use case, overkill (and expensive) if you just need BPM/key.
Not really a music analysis API — it's a music industry intelligence platform. Wrong tool for audio features.
What we all used. Closed to new apps since November 2024; only apps with extended access from before then still get it. Spotify's February 2026 update made even Dev Mode require the app owner to hold an active Premium subscription, capped apps at 5 users each, and capped Client IDs per developer (1 at launch, raised to 25 in July 2026) — there is no replacement from Spotify.
FreqBlog now offers POST /identify — Chromaprint fingerprinting matched against the open AcoustID dataset — at £0.34 per 1k recognitions on Pro: ~12× cheaper than AudD's $5/1k entry rate, ~8× cheaper than their $3.60/1k volume rate. And when the matched recording is in our catalog, you get the full audio analysis in the same response.
A different category — built for apps that need a user's Spotify, Apple Music, and Tidal libraries. Not a competitor for audio analysis. They meter at €0.60 per 1k for the first 5M requests (stepping down to €0.45, then €0.30), with a €500/month minimum — roughly 3× FreqBlog Pro’s per-request rate, and every account pays at least €500.
A free public API for hobby projects. Fine for a weekend prototype; not appropriate for a paying product where uptime, support and a billing relationship matter.
One-on-one comparisons — verifiable facts only, including the honest cases where the other tool is the better fit.
Every music metadata API evaluated on 20+ criteria.
← scroll to compare all APIs →
| Feature | FreqBlog | MeloData | GetSongBPM | Cyanite | Soundcharts |
|---|---|---|---|---|---|
| Lookup Method | |||||
| Name-based (artist + title) | ✓ | ✓ (match, then ISRC) | ✓ | ✗ | Partial |
| ISRC lookup | ✓ | ✓ (required) | ✗ | ✗ | ✓ |
| File upload / analyze | ✓ | ✓ | ✗ | ✓ | ✗ |
| Audio Features | |||||
| BPM / Tempo | ✓ | ✓ | ✓ | ✓ | ✗ |
| Musical Key + Mode | ✓ | ✓ | ✓ | ✓ | ✗ |
| Open Key / Camelot notation | ✓ | ✗ | Partial | ✗ | ✗ |
| Energy (0–1) | ✓ | Partial | ✗ | ✓ | ✗ |
| Loudness (dBFS) | ✓ | ✗ | ✗ | ✗ | ✗ |
| Danceability | ✓ | ✓ | ✗ | ✓ | ✗ |
| Valence / Positivity | ✓ | ✗ | ✗ | ✓ | ✗ |
| Acousticness (approx) | ✓ | ✗ | ✗ | ✓ | ✗ |
| Instrumentalness (30s preview) | ✓ | ✗ | ✗ | ✓ | ✗ |
| Time Signature | ✓ | ✗ | ✗ | ✗ | ✗ |
| Mood & Genre | |||||
| Mood (happy/sad/aggressive…) | ✓ | ✗ | ✗ | ✓ | ✗ |
| Genre detection | ✓ | ✗ | ✗ | ✓ | ✓ |
| Advanced Endpoints | |||||
| Search by BPM range | ✓ /bpm | ✗ | ✗ | ✗ | ✗ |
| Search by key (Camelot/Open) | ✓ /key | ✗ | ✗ | ✗ | ✗ |
| MCP server (AI/LLM integration) | ✓ | ✗ | ✗ | ✗ | ✗ |
| Harmonic transition scoring (A→B mix compatibility) | ✓ | ✗ | ✗ | ✗ | ✗ |
| DJ set ordering (energy arc) | ✓ | ✗ | ✗ | ✗ | ✗ |
| Recommendations + related artists (Spotify drop-in) | ✓ | ✗ | ✗ | ✗ | ✗ |
| Pricing | |||||
| Free tier | ✓ 1,000 req/mo | ✓ 1,000 lookups/mo | ✓ | ✗ | ✗ |
| Entry paid tier | £9.99/mo | $19/mo | Free only | €290/mo + usage | $50/mo |
| Flat-rate monthly pricing | ✓ | ✓ | ✓ | ✗ (per track) | ✓ |
MeloData produces accurate results and now lets you start from a title and artist: you match the name to candidate recordings, check the match confidence, then request features by ISRC. That is a sensible design if you want to review matches before you fetch.
FreqBlog does both steps in one call — /lookup?track=&artist= returns the features directly, the same interface Spotify provided — and also covers what a set-building app needs: Camelot and Open Key, /transition scoring, /setlist ordering and DJ-software export.
GetSongBPM is a well-known free resource that does name-based lookup well. If all you need is BPM and key, it's a reasonable choice. The problem is that it only returns two fields.
Spotify's audio-features returned 13 fields including energy, loudness, danceability, valence, acousticness, instrumentalness, liveness, speechiness and time signature. GetSongBPM returns none of those. If your app used any of those fields, you'll need a different solution.
When a track isn't in our catalog, we resolve it through MusicBrainz to AcousticBrainz, an open (CC0) dataset of pre-analyzed music from community music-information-retrieval research. If neither has it, we analyse the track on demand.
The FreqBlog API uses a two-layer lookup: our own catalogue, then MusicBrainz → AcousticBrainz. When a track is in the catalog (1,950,463+ indexed, 1.2 Million+ fully analysed), you get a response in ~25–40ms with keep-alive (p95 under 100ms). For everything else, the API checks AcousticBrainz via MusicBrainz (mood, mood_vector, genre, valence, and more), and otherwise returns 202 and analyses the track on demand (typically ~15s). Total fallback coverage: 7.5M rows. The mood_vector field (5-axis dimensional mood probabilities) is populated when a match is found — for both fallback and catalog tracks — and is null when no match exists.
No other commercial music API offers this combination of instant lookups, an open-data fallback and on-demand analysis.
No credit card. No ISRC. Same fields Spotify gave you, plus more.
Migrating from Spotify? See the step-by-step migration guide →
Get Free API Key View API Docs