🎛️ Spotify Audio Features Scraper

Get Spotify audio features for any track, restored after the November 2024 deprecation. Tempo (BPM), key, energy, danceability, valence, acousticness and more, in the exact JSON shape developers already know. A drop-in alternative to Spotify's removed audio-features endpoint.

No Spotify developer account needed. No API key, no OAuth. Just paste URLs, IDs, ISRCs, or search keywords and hit Run.

---

⚡ Why this scraper?

In November 2024, Spotify removed public access to the audio-features (and audio-analysis) endpoints, breaking countless apps, playlists tools, and ML pipelines. This scraper restores that access: same fields, same value ranges, no migration headache.

• 🎛️ Full audio-features object: tempo, key, mode, time signature, energy, danceability, valence, acousticness, instrumentalness, speechiness, liveness, loudness, duration.
• 🔌 Drop-in shape: returns the exact audio_features JSON structure Spotify used to serve. Swap the source, keep your code.
• 🧠 ML embedding (optional): a normalized 12-feature vector per track, ready for similarity search, clustering, and recommendation models.
• 🔀 Flexible input: Spotify URLs, track IDs, ISRCs, or plain search keywords, mix and match freely.
• 📦 Clean, bulk output: one row per track as JSON, CSV, or Excel, no rate-limit juggling.

---

🏆 How it compares

Feature	This scraper	Spotify Web API	Other scrapers
🎛️ Audio features	✅	❌ Deprecated (Nov 2024)	❌ Most can't
🔑 API key / OAuth required	❌ None	✅ Required	⚠️ Varies
🏷️ ISRC input	✅	❌	❌
🔍 Search by keyword	✅	✅	❌ Most don't
🧠 ML embedding vector	✅	❌	❌
📦 Bulk + JSON/CSV/Excel	✅	⚠️ Manual	⚠️ Varies

---

📊 What data do you get?

For every track, the full audio-features object plus a minimal track identity:

Feature	Description	Example
Tempo	BPM (beats per minute)	122.746
Key	Musical key (0 = C, 1 = C#/Db, ..., 11 = B)	2
Mode	Major (1) or Minor (0)	1
Time Signature	Estimated time signature	4
Energy	Intensity and activity (0.0 to 1.0)	0.697
Danceability	How suitable for dancing (0.0 to 1.0)	0.613
Valence	Musical positivity / mood (0.0 to 1.0)	0.476
Acousticness	Confidence the track is acoustic (0.0 to 1.0)	0.0194
Instrumentalness	Likelihood of no vocals (0.0 to 1.0)	0.0
Speechiness	Presence of spoken words (0.0 to 1.0)	0.133
Liveness	Probability of a live recording (0.0 to 1.0)	0.332
Loudness	Overall loudness in decibels (dB)	-8.618
Duration	Track length in milliseconds	320357

Each result also includes the track's name, Spotify ID, ISRC, popularity, cover art, URL and URI, plus the standard id, uri, track_href, analysis_url and type fields, so it parses exactly like Spotify's original response.

🧠 ML Embedding (optional)

Enable Include ML Embedding Vector to add a normalized 12-feature vector per track:

"embedding": {
  "id": "0DiWol3AO6WpXZgp0goxAV",
  "features": ["danceability", "energy", "acousticness", "instrumentalness", "liveness", "speechiness", "valence", "loudness", "tempo", "key", "mode", "time_signature"],
  "vector": [0.613, 0.697, 0.0194, 0.0, 0.332, 0.133, 0.476, 0.856367, 0.490984, 0.181818, 1.0, 0.571429]
}

---

🚀 How to use

1. Enter your tracks: Paste Spotify URLs, track IDs, ISRCs, or keywords in the Tracks field.
2. Toggle the embedding: Enable the ML vector if you need it for similarity/ML work.
3. Run: Click Start and get audio features back in seconds.

⚙️ Options

Option	Default	Description
🧠 Include ML Embedding Vector	⬜ Off	Add a normalized 12-feature vector per track
🔢 Max results per keyword	10	Cap the number of results per search keyword (URL/ID/ISRC always return one)

Input examples

You can mix input types freely:

https://open.spotify.com/track/0DiWol3AO6WpXZgp0goxAV
49X0LAl6faAusYq02PRAY6
GBDUW0000053
Music Sounds Better With You

• 🔗 URL: Any open.spotify.com/track/ link
• 🆔 ID: The 22-character Spotify track ID
• 🏷️ ISRC: International Standard Recording Code (e.g. GBDUW0000053)
• 🔍 Keyword: A search term returning up to 50 results (configurable)

---

📤 Output

Each result is a JSON object. Results are pushed to the dataset and saved as individual JSON files in the key-value store (e.g. track_0DiWol3AO6WpXZgp0goxAV.json).

{
    "name": "One More Time",
    "type": "track",
    "url": "https://open.spotify.com/track/0DiWol3AO6WpXZgp0goxAV",
    "image": "https://i.scdn.co/image/ab67616d0000b273...",
    "success": true,
    "result": "1/1",
    "mode": "url",
    "target": "https://open.spotify.com/track/0DiWol3AO6WpXZgp0goxAV",
    "error": null,
    "track": {
        "track_image": "https://i.scdn.co/image/ab67616d0000b273...",
        "track_name": "One More Time",
        "track_id": "0DiWol3AO6WpXZgp0goxAV",
        "track_url": "https://open.spotify.com/track/0DiWol3AO6WpXZgp0goxAV",
        "track_uri": "spotify:track:0DiWol3AO6WpXZgp0goxAV",
        "track_isrc": "GBDUW0000053",
        "track_popularity": 80
    },
    "audio_features": {
        "acousticness": 0.0194,
        "danceability": 0.613,
        "energy": 0.697,
        "instrumentalness": 0.0,
        "liveness": 0.332,
        "speechiness": 0.133,
        "valence": 0.476,
        "loudness": -8.618,
        "tempo": 122.746,
        "key": 2,
        "mode": 1,
        "time_signature": 4,
        "duration_ms": 320357,
        "id": "0DiWol3AO6WpXZgp0goxAV",
        "uri": "spotify:track:0DiWol3AO6WpXZgp0goxAV",
        "track_href": "https://api.spotify.com/v1/tracks/0DiWol3AO6WpXZgp0goxAV",
        "analysis_url": "https://api.spotify.com/v1/audio-analysis/0DiWol3AO6WpXZgp0goxAV",
        "type": "audio_features"
    }
}

Available as JSON, CSV, Excel, or via the Apify API and dataset exports.

---

💡 Use cases

• 🤖 ML & recommendation pipelines: Feed normalized audio features (or the embedding vector) straight into your models.
• 📊 Music analytics: Compare tempo, energy and valence distributions across catalogs, genres, or playlists.
• 🎶 Playlist & mood tooling: Build energy- or mood-based playlists from valence/energy/danceability.
• 📈 Trend research: Track how the sonic profile of releases shifts over time.
• 🛠️ App migration: Restore the audio-features data your app lost when Spotify deprecated the endpoint.
• 🎓 Academic research: Collect structured audio-feature datasets for musicology or computational analysis.

---

❓ FAQ

Do I need a Spotify account or API key? No. No developer account, no token, no OAuth, just paste your input and Run.

Spotify deprecated audio features in Nov 2024, how does this still work? This scraper restores access to the same audio-feature data and returns it in Spotify's original audio_features JSON shape, so existing code keeps working.

What can I use as input? Spotify track URLs, 22-character track IDs, ISRCs, or plain search keywords. Mix all four freely in one run.

Is the output the same shape as Spotify's old endpoint? Yes. The audio_features object uses the same field names and value ranges, so it's a drop-in replacement.

What's the ML embedding for? It's a normalized, fixed-order 12-feature vector per track, convenient for similarity search, clustering, and feeding recommendation/ML models without re-normalizing.

How many tracks can I process in one run? As many as you like. Search keywords return up to 50 results each (configurable); URL/ID/ISRC inputs return exactly one each.

What formats can I export? JSON, CSV, Excel, or via the Apify API and dataset exports.

---

🔌 Integrations

Connect results to your existing workflow:

• 📗 Google Sheets: Auto-export results to a spreadsheet
• 🪝 Webhooks: Trigger downstream actions when a run completes
• ⚡ Zapier / Make: Plug into 1000+ apps
• 🛠️ API: Fetch results programmatically from any language
• 🔄 Scheduled runs: Refresh your audio-feature datasets on a schedule