Introduction
# AudioPod AI
Full audio processing API: music generation, stem separation, TTS, noise reduction, transcription, speaker separation, wallet management.
## Setup
```bash pip install audiopod # Python npm install audiopod # Node.js ```
Auth: set `AUDIOPOD_API_KEY` env var or pass to client constructor.
### Getting an API Key 1. Sign up at https://audiopod.ai/auth/signup (free, no credit card required) 2. Go to https://www.audiopod.ai/dashboard/account/api-keys 3. Click "Create API Key" and copy the key (starts with `ap_`) 4. Add funds to your wallet at https://www.audiopod.ai/dashboard/account/wallet (pay-as-you-go, no subscription)
```python from audiopod import AudioPod client = AudioPod() # uses AUDIOPOD_API_KEY env var # or: client = AudioPod(api_key="ap_...") ```
---
## AI Music Generation
Generate songs, rap, instrumentals, samples, and vocals from text prompts.
**Tasks:** `text2music` (song with vocals), `text2rap` (rap), `prompt2instrumental` (instrumental), `lyric2vocals` (vocals only), `text2samples` (loops/samples), `audio2audio` (style transfer), `songbloom`
### Python SDK
```python # Generate a full song with lyrics result = client.music.song( prompt="Upbeat pop, synth, drums, 120 bpm, female vocals, radio-ready", lyrics="Verse 1:\nWalking down the street on a sunny day\n\nChorus:\nWe're on fire tonight!", duration=60 ) print(result["output_url"])
# Generate rap result = client.music.rap( prompt="Lo-Fi Hip Hop, 100 BPM, male rap, melancholy, keyboard chords", lyrics="Verse 1:\nStarted from the bottom, now we climbing...", duration=60 )
# Generate instrumental (no lyrics needed) result = client.music.instrumental( prompt="Atmospheric ambient soundscape, uplifting, driving mood", duration=30 )
# Generic generate with explicit task result = client.music.generate( prompt="Electronic dance music, high energy", task="text2samples", # any task type duration=30 )
# Async: submit then poll job = client.music.create( prompt="Chill lofi beat", duration=30, task="prompt2instrumental" ) result = client.music.wait_for_completion(job["id"], timeout=600)
# Get available genre presets presets = client.music.get_presets()
# List/manage jobs jobs = client.music.list(skip=0, limit=50) job = client.music.get(job_id=123) client.music.delete(job_id=123) ```
### cURL
```bash # Song with lyrics curl -X POST "https://api.audiopod.ai/api/v1/music/text2music" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"prompt":"upbeat pop, synth, 120bpm, female vocals", "lyrics":"Walking down the street...", "audio_duration":60}'
# Rap curl -X POST "https://api.audiopod.ai/api/v1/music/text2rap" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"prompt":"Lo-Fi Hip Hop, male rap, 100 BPM", "lyrics":"Started from the bottom...", "audio_duration":60}'
# Instrumental curl -X POST "https://api.audiopod.ai/api/v1/music/prompt2instrumental" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"prompt":"ambient soundscape, uplifting", "audio_duration":30}'
# Samples/loops curl -X POST "https://api.audiopod.ai/api/v1/music/text2samples" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"prompt":"drum loop, sad mood", "audio_duration":15}'
# Vocals only curl -X POST "https://api.audiopod.ai/api/v1/music/lyric2vocals" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"prompt":"clean vocals, happy", "lyrics":"Eternal chorus of unity...", "audio_duration":30}'
# Check job status / get result curl "https://api.audiopod.ai/api/v1/music/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Get genre presets curl "https://api.audiopod.ai/api/v1/music/presets" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# List jobs curl "https://api.audiopod.ai/api/v1/music/jobs?skip=0&limit=50" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Delete job curl -X DELETE "https://api.audiopod.ai/api/v1/music/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY" ```
### Parameters
| Field | Required | Description | |-------|----------|-------------| | prompt | yes | Style/genre description | | lyrics | for song/rap/vocals | Song lyrics with verse/chorus structure | | audio_duration | no | Duration in seconds (default: 30) | | genre_preset | no | Genre preset name (from presets endpoint) | | display_name | no | Track display name |
---
## Stem Separation
Split audio into individual instrument/vocal tracks.
### Modes
| Mode | Stems | Output | Use Case | |------|-------|--------|----------| | single | 1 | Specified stem only | Vocal isolation, drum extraction | | two | 2 | vocals + instrumental | Karaoke tracks | | four | 4 | vocals, drums, bass, other | Standard remixing (default) | | six | 6 | + guitar, piano | Full instrument separation | | producer | 8 | + kick, snare, hihat | Beat production | | studio | 12 | + cymbals, sub_bass, synth | Professional mixing | | mastering | 16 | Maximum detail | Forensic analysis |
**Single stem options:** vocals, drums, bass, guitar, piano, other
### Python SDK
```python # Sync: extract and wait for result result = client.stems.separate( url="https://youtube.com/watch?v=VIDEO_ID", mode="six", timeout=600 ) for stem, url in result["download_urls"].items(): print(f"{stem}: {url}")
# From local file result = client.stems.separate(file="/path/to/song.mp3", mode="four")
# Single stem extraction result = client.stems.separate( url="https://youtube.com/watch?v=ID", mode="single", stem="vocals" )
# Async: submit then poll job = client.stems.extract(url="https://youtube.com/watch?v=ID", mode="six") print(f"Job ID: {job['id']}") status = client.stems.status(job["id"]) # or wait: result = client.stems.wait_for_completion(job["id"], timeout=600)
# List available modes modes = client.stems.modes()
# Job management jobs = client.stems.list(skip=0, limit=50, status="COMPLETED") job = client.stems.get(job_id=1234) client.stems.delete(job_id=1234) ```
### cURL
```bash # Extract from URL curl -X POST "https://api.audiopod.ai/api/v1/stem-extraction/api/extract" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "url=https://youtube.com/watch?v=VIDEO_ID" \ -F "mode=six"
# Extract from file curl -X POST "https://api.audiopod.ai/api/v1/stem-extraction/api/extract" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "file=@/path/to/song.mp3" \ -F "mode=four"
# Single stem curl -X POST "https://api.audiopod.ai/api/v1/stem-extraction/api/extract" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "url=URL" \ -F "mode=single" \ -F "stem=vocals"
# Check job status curl "https://api.audiopod.ai/api/v1/stem-extraction/status/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# List available modes curl "https://api.audiopod.ai/api/v1/stem-extraction/modes" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# List jobs (filter by status: PENDING, PROCESSING, COMPLETED, FAILED) curl "https://api.audiopod.ai/api/v1/stem-extraction/jobs?skip=0&limit=50&status=COMPLETED" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Get specific job curl "https://api.audiopod.ai/api/v1/stem-extraction/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Delete job curl -X DELETE "https://api.audiopod.ai/api/v1/stem-extraction/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY" ```
### Response Format
```json { "id": 1234, "status": "COMPLETED", "download_urls": { "vocals": "https://...", "drums": "https://...", "bass": "https://...", "other": "https://..." }, "quality_scores": { "vocals": 0.95, "drums": 0.88 } } ```
---
## Text to Speech
Generate speech from text with 50+ voices in 60+ languages. Supports voice cloning.
### Voice Types
- **50+ production-ready voices** — multilingual, supporting 60+ languages with auto-detection - **Custom clones** — clone any voice with ~5 seconds of audio sample
### Python SDK
```python # Generate speech and wait for result result = client.voice.generate( text="Hello, world! This is a test.", voice_id=123, speed=1.0 ) print(result["output_url"])
# Async: submit then poll job = client.voice.speak( text="Hello world", voice_id=123, speed=1.0 ) status = client.voice.get_job(job["id"]) result = client.voice.wait_for_completion(job["id"], timeout=300)
# List all available voices voices = client.voice.list() for v in voices: print(f"{v['id']}: {v['name']}")
# Clone a voice (needs ~5 sec audio sample) new_voice = client.voice.create( name="My Voice Clone", audio_file="./sample.mp3", description="Cloned from recording" )
# Get/delete voice voice = client.voice.get(voice_id=123) client.voice.delete(voice_id=123) ```
### cURL (Raw HTTP — most reliable)
```bash # List all voices curl "https://api.audiopod.ai/api/v1/voice/voice-profiles" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Generate speech (FORM DATA, not JSON!) curl -X POST "https://api.audiopod.ai/api/v1/voice/voices/{VOICE_UUID}/generate" \ -H "Authorization: Bearer $AUDIOPOD_API_KEY" \ -d "input_text=Hello world, this is a test" \ -d "audio_format=mp3" \ -d "speed=1.0"
# Poll job status curl "https://api.audiopod.ai/api/v1/voice/tts-jobs/{JOB_ID}/status" \ -H "Authorization: Bearer $AUDIOPOD_API_KEY"
# SDK-style endpoints (alternative) # Generate via SDK endpoint curl -X POST "https://api.audiopod.ai/api/v1/voice/tts/generate" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"text":"Hello world","voice_id":123,"speed":1.0}'
# Poll via SDK endpoint curl "https://api.audiopod.ai/api/v1/voice/tts/status/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# List voices (SDK endpoint) curl "https://api.audiopod.ai/api/v1/voice/voices" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Clone a voice curl -X POST "https://api.audiopod.ai/api/v1/voice/voices" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "name=My Voice" \ -F "[email protected]" \ -F "description=Cloned voice"
# Delete voice curl -X DELETE "https://api.audiopod.ai/api/v1/voice/voices/VOICE_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY" ```
### Generate Parameters
| Field | Required | Description | |-------|----------|-------------| | input_text | yes | Text to speak (max 5000 chars). Use `input_text` for raw HTTP, `text` for SDK | | audio_format | no | mp3, wav, ogg (default: mp3) | | speed | no | 0.25 - 4.0 (default: 1.0) | | language | no | ISO code, auto-detected if omitted |
### Response Format
```json // Generate response {"job_id": 12345, "status": "pending", "credits_reserved": 25}
// Status response (completed) {"status": "completed", "output_url": "https://r2-url/generated.mp3"} ```
### Important Notes
- Raw HTTP generate endpoint uses **form data**, not JSON. Field is `input_text` not `text` - SDK endpoint (`/api/v1/voice/tts/generate`) uses JSON with field `text` - Output files may be WAV disguised as .mp3 — convert with `ffmpeg -i output.mp3 -c:a aac real.m4a` - ~55 credits per generation, wallet-based billing
---
## Speaker Separation
Separate audio by speaker with automatic diarization.
### Python SDK
```python # Diarize and wait for result result = client.speaker.identify( file="./meeting.mp3", num_speakers=3, # optional hint for accuracy timeout=600 ) for segment in result["segments"]: print(f"Speaker {segment['speaker']}: {segment['text']} [{segment['start']:.1f}s - {segment['end']:.1f}s]")
# From URL result = client.speaker.identify( url="https://youtube.com/watch?v=VIDEO_ID", num_speakers=2 )
# Async: submit then poll job = client.speaker.diarize( file="./meeting.mp3", num_speakers=3 ) result = client.speaker.wait_for_completion(job["id"], timeout=600)
# Job management jobs = client.speaker.list(skip=0, limit=50, status="COMPLETED") job = client.speaker.get(job_id=123) client.speaker.delete(job_id=123) ```
### cURL
```bash # Diarize from file curl -X POST "https://api.audiopod.ai/api/v1/speaker/diarize" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "[email protected]" \ -F "num_speakers=3"
# Diarize from URL curl -X POST "https://api.audiopod.ai/api/v1/speaker/diarize" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "url=https://youtube.com/watch?v=VIDEO_ID" \ -F "num_speakers=2"
# Check job status curl "https://api.audiopod.ai/api/v1/speaker/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# List jobs curl "https://api.audiopod.ai/api/v1/speaker/jobs?skip=0&limit=50" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Delete job curl -X DELETE "https://api.audiopod.ai/api/v1/speaker/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY" ```
---
## Speech to Text (Transcription)
Transcribe audio/video with speaker diarization, word-level timestamps, and multiple output formats.
### Python SDK
```python # Transcribe URL and wait result = client.transcription.transcribe( url="https://youtube.com/watch?v=VIDEO_ID", speaker_diarization=True, min_speakers=2, max_speakers=5, timeout=600 ) print(f"Language: {result['detected_language']}") for seg in result["segments"]: print(f"[{seg['start']:.1f}s] {seg.get('speaker','?')}: {seg['text']}")
# Batch: multiple URLs at once result = client.transcription.transcribe( urls=["https://youtube.com/watch?v=ID1", "https://youtube.com/watch?v=ID2"], speaker_diarization=True )
# Upload local file job = client.transcription.upload( file_path="./recording.mp3", language="en", speaker_diarization=True ) result = client.transcription.wait_for_completion(job["id"], timeout=600)
# Async: submit then poll job = client.transcription.create( url="https://youtube.com/watch?v=ID", language="en", speaker_diarization=True, word_timestamps=True, min_speakers=2, max_speakers=4 ) result = client.transcription.wait_for_completion(job["id"], timeout=600)
# Get transcript in different formats transcript_json = client.transcription.get_transcript(job_id=123, format="json") transcript_srt = client.transcription.get_transcript(job_id=123, format="srt") transcript_vtt = client.transcription.get_transcript(job_id=123, format="vtt") transcript_txt = client.transcription.get_transcript(job_id=123, format="txt")
# Job management jobs = client.transcription.list(skip=0, limit=50, status="COMPLETED") job = client.transcription.get(job_id=123) client.transcription.delete(job_id=123) ```
### cURL
```bash # Transcribe from URL curl -X POST "https://api.audiopod.ai/api/v1/transcribe/transcribe" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"url":"https://youtube.com/watch?v=ID","enable_speaker_diarization":true,"word_timestamps":true}'
# Transcribe multiple URLs curl -X POST "https://api.audiopod.ai/api/v1/transcribe/transcribe" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"urls":["URL1","URL2"],"enable_speaker_diarization":true}'
# Upload file for transcription curl -X POST "https://api.audiopod.ai/api/v1/transcribe/transcribe-upload" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "[email protected]" \ -F "language=en" \ -F "enable_speaker_diarization=true"
# Get job status curl "https://api.audiopod.ai/api/v1/transcribe/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Get transcript in specific format (json, srt, vtt, txt) curl "https://api.audiopod.ai/api/v1/transcribe/jobs/JOB_ID/transcript?format=srt" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# List jobs curl "https://api.audiopod.ai/api/v1/transcribe/jobs?offset=0&limit=50" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Delete job curl -X DELETE "https://api.audiopod.ai/api/v1/transcribe/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY" ```
### Parameters
| Field | Required | Description | |-------|----------|-------------| | url / urls | yes (or file) | URL(s) to transcribe (YouTube, SoundCloud, direct links) | | language | no | ISO 639-1 code (auto-detected if omitted) | | enable_speaker_diarization | no | Enable speaker identification (default: false) | | min_speakers / max_speakers | no | Speaker count hints for better diarization | | word_timestamps | no | Enable word-level timestamps (default: true) |
### Output Formats
- **json** — Full structured output with segments, timestamps, speakers - **srt** — SubRip subtitle format - **vtt** — WebVTT subtitle format - **txt** — Plain text transcript
---
## Noise Reduction
Remove background noise from audio/video files.
### Python SDK
```python # Denoise and wait for result result = client.denoiser.denoise(file="./noisy-audio.mp3", timeout=600) print(f"Clean audio: {result['output_url']}")
# From URL result = client.denoiser.denoise(url="https://example.com/noisy.mp3")
# Async: submit then poll job = client.denoiser.create(file="./noisy-audio.mp3") result = client.denoiser.wait_for_completion(job["id"], timeout=600)
# From URL (async) job = client.denoiser.create(url="https://example.com/noisy.mp3")
# Job management jobs = client.denoiser.list(skip=0, limit=50, status="COMPLETED") job = client.denoiser.get(job_id=123) client.denoiser.delete(job_id=123) ```
### cURL
```bash # Denoise from file curl -X POST "https://api.audiopod.ai/api/v1/denoiser/denoise" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "[email protected]"
# Denoise from URL curl -X POST "https://api.audiopod.ai/api/v1/denoiser/denoise" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -F "url=https://example.com/noisy.mp3"
# Check job status curl "https://api.audiopod.ai/api/v1/denoiser/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# List jobs curl "https://api.audiopod.ai/api/v1/denoiser/jobs?skip=0&limit=50" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Delete job curl -X DELETE "https://api.audiopod.ai/api/v1/denoiser/jobs/JOB_ID" \ -H "X-API-Key: $AUDIOPOD_API_KEY" ```
---
## Wallet & Billing
Check balance, estimate costs, and view usage history.
### Python SDK
```python # Get current balance balance = client.wallet.get_balance() print(f"Balance: ${balance['balance_usd']}")
# Check if balance is sufficient for an operation check = client.wallet.check_balance( service_type="stem_extraction", duration_seconds=180 ) print(f"Sufficient: {check['sufficient']}")
# Estimate cost before running estimate = client.wallet.estimate_cost( service_type="transcription", duration_seconds=300 ) print(f"Cost: ${estimate['cost_usd']}")
# Get pricing for all services pricing = client.wallet.get_pricing()
# View usage history usage = client.wallet.get_usage(page=1, limit=50) ```
### cURL
```bash # Get balance curl "https://api.audiopod.ai/api/v1/api-wallet/balance" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Check balance sufficiency curl -X POST "https://api.audiopod.ai/api/v1/api-wallet/check-balance" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"service_type":"stem_extraction","duration_seconds":180}'
# Estimate cost curl -X POST "https://api.audiopod.ai/api/v1/api-wallet/estimate-cost" \ -H "X-API-Key: $AUDIOPOD_API_KEY" \ -H "Content-Type: application/json" \ -d '{"service_type":"transcription","duration_seconds":300}'
# Get pricing curl "https://api.audiopod.ai/api/v1/api-wallet/pricing" \ -H "X-API-Key: $AUDIOPOD_API_KEY"
# Usage history curl "https://api.audiopod.ai/api/v1/api-wallet/usage?page=1&limit=50" \ -H "X-API-Key: $AUDIOPOD_API_KEY" ```
---
## API Endpoint Summary
| Service | Endpoint | Method | |---------|----------|--------| | **Music** | `/api/v1/music/{task}` | POST | | Music jobs | `/api/v1/music/jobs/{id}` | GET/DELETE | | Music presets | `/api/v1/music/presets` | GET | | **Stems** | `/api/v1/stem-extraction/api/extract` | POST (multipart) | | Stems status | `/api/v1/stem-extraction/status/{id}` | GET | | Stems modes | `/api/v1/stem-extraction/modes` | GET | | Stems jobs | `/api/v1/stem-extraction/jobs` | GET | | **TTS** generate | `/api/v1/voice/voices/{uuid}/generate` | POST (form data) | | TTS generate (SDK) | `/api/v1/voice/tts/generate` | POST (JSON) | | TTS status | `/api/v1/voice/tts-jobs/{id}/status` | GET | | TTS status (SDK) | `/api/v1/voice/tts/status/{id}` | GET | | Voice list | `/api/v1/voice/voice-profiles` | GET | | Voice list (SDK) | `/api/v1/voice/voices` | GET | | **Speaker** | `/api/v1/speaker/diarize` | POST (multipart) | | Speaker jobs | `/api/v1/speaker/jobs/{id}` | GET/DELETE | | **Transcribe** URL | `/api/v1/transcribe/transcribe` | POST (JSON) | | Transcribe upload | `/api/v1/transcribe/transcribe-upload` | POST (multipart) | | Transcript output | `/api/v1/transcribe/jobs/{id}/transcript?format=` | GET | | Transcribe jobs | `/api/v1/transcribe/jobs` | GET | | **Denoise** | `/api/v1/denoiser/denoise` | POST (multipart) | | Denoise jobs | `/api/v1/denoiser/jobs/{id}` | GET/DELETE | | **Wallet** balance | `/api/v1/api-wallet/balance` | GET | | Wallet pricing | `/api/v1/api-wallet/pricing` | GET | | Wallet usage | `/api/v1/api-wallet/usage` | GET |
## Auth Headers
Two auth styles work: - `X-API-Key: ap_...` — works for most endpoints - `Authorization: Bearer ap_...` — works for TTS generate/status
## Known Issues
- SDK method signatures may differ from raw API — when in doubt, use cURL examples - TTS output stored on Cloudflare R2, download via `output_url` in job status - TTS output files may be WAV disguised as .mp3 — convert with ffmpeg before sending via WhatsApp