---
title: "Audio API"
description: "Transcribe audio with Whisper, generate speech with TTS, and create music from text prompts."
canonical: "https://docs.ainative.studio/docs/api/audio"
last-updated: "2026-10-26T19:08:16.785Z"
---

# Audio API

Source: https://docs.ainative.studio/docs/api/audio

> Transcribe audio with Whisper, generate speech with TTS, and create music from text prompts.

# Audio API

Transcribe audio files, generate speech, and create music. All endpoints consume credits and require authentication.

**Base URL:** `https://api.ainative.studio/api/v1/audio`

---

## Transcription (Whisper)

Transcribe audio files to text using OpenAI Whisper.

<APIEndpoint method="POST" path="/api/v1/audio/transcriptions" auth />

**Cost:** ~6 credits/minute of audio | **Max file size:** 25 MB

```bash
curl -X POST https://api.ainative.studio/api/v1/audio/transcriptions \
  -H "Authorization: Bearer $TOKEN" \
  -F file=@recording.mp3 \
  -F model=whisper-1 \
  -F response_format=json
```

```python
import requests

with open("recording.mp3", "rb") as f:
    response = requests.post(
        "https://api.ainative.studio/api/v1/audio/transcriptions",
        headers={"Authorization": f"Bearer {token}"},
        files={"file": f},
        data={"model": "whisper-1", "response_format": "json"}
    )

print(response.json()["text"])
```

**Parameters (form data):**

| Parameter | Type | Required | Default | Description |
|-----------|------|----------|---------|-------------|
| `file` | `file` | Yes | — | Audio file (mp3, mp4, wav, webm, m4a, mpeg, mpga) |
| `model` | `string` | No | `whisper-1` | Whisper model |
| `language` | `string` | No | auto-detect | ISO-639-1 language code |
| `prompt` | `string` | No | — | Hint text to guide transcription |
| `response_format` | `string` | No | `text` | `text`, `json`, `srt`, or `vtt` |
| `temperature` | `float` | No | — | Sampling temperature (0.0-1.0) |

**Response:**

| Field | Type | Description |
|-------|------|-------------|
| `text` | `string` | Transcribed text |
| `cost_credits` | `int` | Credits charged |
| `duration_seconds` | `float` | Audio duration |
| `language` | `string` | Detected language |

---

## Translation (Whisper)

Translate audio from any language to English.

<APIEndpoint method="POST" path="/api/v1/audio/translations" auth />

Same parameters and cost as transcription. Input can be any supported language; output is always English.

---

## Text-to-Speech

Generate speech audio from text.

<APIEndpoint method="POST" path="/api/v1/audio/tts" auth />

```bash
curl -X POST https://api.ainative.studio/api/v1/audio/tts \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Hello from AINative Studio",
    "model_id": "facebook/mms-tts-eng"
  }'
```

**Parameters:**

| Parameter | Type | Required | Default | Description |
|-----------|------|----------|---------|-------------|
| `text` | `string` | Yes | — | Text to synthesize (1-5000 chars) |
| `model_id` | `string` | No | `facebook/mms-tts-eng` | TTS model ID |
| `voice` | `string` | No | — | Voice/speaker parameter |

**Response:**

| Field | Type | Description |
|-------|------|-------------|
| `audio_base64` | `string` | Base64-encoded audio |
| `duration_seconds` | `float` | Audio length |
| `cost_credits` | `int` | Credits charged |
| `format` | `string` | Audio format (wav/mp3) |

:::tip Two TTS Endpoints
- **`/api/v1/audio/tts`** — HuggingFace models (free tier, lower quality)
- **`/api/v1/multimodal/tts`** — MiniMax TTS (14 credits, higher quality)

Choose based on your quality and budget requirements.
:::

---

## Music Generation

Generate music from text descriptions.

<APIEndpoint method="POST" path="/api/v1/audio/music" auth />

```bash
curl -X POST https://api.ainative.studio/api/v1/audio/music \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Upbeat electronic music with synthesizer melody",
    "duration": 15
  }'
```

**Parameters:**

| Parameter | Type | Required | Default | Description |
|-----------|------|----------|---------|-------------|
| `prompt` | `string` | Yes | — | Music description (1-500 chars) |
| `model_id` | `string` | No | `minimax-music-v2` | Music model |
| `duration` | `int` | No | `10` | Duration in seconds (5-60) |

**Response:**

| Field | Type | Description |
|-------|------|-------------|
| `audio_base64` | `string` | Base64-encoded audio |
| `duration_seconds` | `float` | Actual duration |
| `cost_credits` | `int` | Credits charged |
| `format` | `string` | Audio format |
| `prompt` | `string` | Prompt used |

---

## For AI Agents

- Use **transcription** to process voice memos, meeting recordings, or audio commands
- Use **TTS** to generate voice responses for users or narrate content
- Use **music generation** for background audio in video content
- Audio files can be stored in ZeroDB file storage for later retrieval