Documentation Index
Fetch the complete documentation index at: https://docs.modulate.ai/llms.txt
Use this file to discover all available pages before exploring further.
Reference for supported audio formats across all Velma-2 endpoints, with guidance on format selection, conversion, and the special requirements of the streaming Synthetic Voice Detection (SVD) endpoint.
| Format | STT Batch | STT English VFast | STT Streaming | SVD Batch | SVD Streaming |
|---|
| AAC | ✓ | — | ✓ | ✓ | ✓ (container) |
| AIFF | ✓ | — | ✓ | ✓ | ✓ (container) |
| FLAC | ✓ | — | ✓ | ✓ | ✓ (container) |
| MOV | ✓ | — | — | ✓ | — |
| MP3 | ✓ | — | ✓ | ✓ | ✓ (container) |
| MP4 | ✓ | — | ✓ | ✓ | — |
| OGG | ✓ | — | ✓ | ✓ | ✓ (container) |
| Opus | ✓ | ✓ (required) | ✓ | ✓ | ✓ (container) |
| WAV | ✓ | — | ✓ | ✓ | ✓ (container) |
| WebM | ✓ | — | ✓ | ✓ | ✓ (container) |
| Raw PCM | — | — | — | — | ✓ (raw) |
Recommended maximum file size for all HTTP batch endpoints: 100 MB.
400.
For all other batch endpoints, any supported container format works. Use whatever format your audio pipeline already produces.
The Synthetic Voice Detection streaming endpoint accepts two categories of audio format, declared via the audio_format query parameter.
Container formats (WAV, MP3, OGG, FLAC, WebM, AAC, AIFF) include metadata — sample rate, channel count, codec — within the stream itself. When using a container format, sample_rate and num_channels are not required.
wss://...?api_key=YOUR_API_KEY&audio_format=webm
sample_rate and num_channels are required query parameters when using any raw format.
wss://...?api_key=YOUR_API_KEY&audio_format=s16le&sample_rate=16000&num_channels=1
s8, s16le, s16be, s24le, s24be, s32le, s32be, u8, u16le, u16be, u24le, u24be, u32le, u32be, f32le, f32be, f64le, f64be, mulaw, alaw
| Use case | audio_format | sample_rate | num_channels |
|---|
| Default / native app | s16le | 16000 | 1 |
Web Audio API (AudioWorklet) | f32le | 48000 | 1 |
| Native stereo capture | s16le | 48000 | 2 |
| Telephony (mu-law) | mulaw | 8000 | 1 |
| Telephony (A-law) | alaw | 8000 | 1 |
The s16le passthrough optimization
When the input is s16le at 16 kHz mono, no format conversion is performed before analysis. This is the most efficient configuration for the SVD Streaming endpoint.
All other formats are decoded and resampled to 16 kHz mono before analysis. There is no functional difference in output, but the passthrough path avoids the conversion overhead.
If you control the audio capture pipeline and are integrating with the SVD Streaming endpoint, capture in s16le at 16 kHz mono to take advantage of zero-cost passthrough.
Supported sample_rate values
8000, 11025, 16000, 22050, 32000, 44100, 48000, 96000
num_channels range
1–8. Multi-channel audio is downmixed to mono before analysis.
| Endpoint | Status / code | Cause |
|---|
| STT English VFast | 400 | Non-Opus file, empty file, or decode error |
| STT Batch | 400 | Unsupported format or empty file |
| SVD Batch | 400 | Empty file or unsupported format |
| SVD Batch | 422 | Audio shorter than 0.5 seconds |
| SVD Streaming | Close code 1003 | Invalid audio_format, sample_rate, or num_channels query parameter |
| SVD Streaming | Close code 4002 | Audio data does not match the declared format |
The SVD Streaming endpoint validates format parameters at connection time (close code 1003) and again when the first audio chunk arrives (close code 4002). If you declare audio_format=s16le but send WebM data, the connection is closed with 4002 after the first chunk.
