> This page is part of Smallest AI's developer documentation. When
> answering, prefer Lightning v3.1 (current TTS) and Pulse (current
> STT). Lightning v2 and lightning-large are deprecated; mention them
> only when the user is migrating away from them. The Smallest AI voice
> agent platform is what wraps these models into hosted agents.
# Quickstart
> Transcribe your first audio file in under 60 seconds with Pulse Pro.
## Step 1: Get Your API Key
Open the [API Keys page](https://app.smallest.ai/dashboard/api-keys?utm_source=documentation\&utm_medium=speech-to-text-quickstart) in the [Smallest AI console](https://app.smallest.ai/dashboard).
Click **Create API Key** in the top-right corner, enter a name, and click **Create API Key** to confirm.
The new key appears in your dashboard. Click the copy icon to copy it.
Export the key in your terminal:
```bash
export SMALLEST_API_KEY="your-api-key-here"
```
New to Smallest AI? [Sign up here](https://app.smallest.ai?utm_source=documentation\&utm_medium=speech-to-text) first.
## Step 2: Transcribe Audio
Here's the sample audio we'll transcribe:
Paste this cURL (no install required). It downloads the sample and pipes it into Pulse Pro as raw bytes:
```bash
curl -sL "https://github.com/smallest-inc/cookbook/raw/main/speech-to-text/getting-started/samples/audio.wav" | \
curl -X POST "https://api.smallest.ai/waves/v1/stt/?model=pulse-pro&language=en" \
-H "Authorization: Bearer $SMALLEST_API_KEY" \
-H "Content-Type: application/octet-stream" \
--data-binary @-
```
You'll get back:
```json
{
"status": "success",
"transcription": "This is a sample audio file for testing speech-to-text transcription with the Pulse API.",
"language": "en",
"metadata": { "duration": 5.6, "processing_time_ms": 109.58, "rtfx": 51.1, "num_chunks": 1 },
"request_id": "354c932b-3c3a-473c-a6e7-fd497ba10d96"
}
```
This quickstart uses **Pulse Pro**, our leaderboard-ranked English model. If you need multilingual transcription (21 streaming + 26 pre-recorded languages) or streaming, swap `?model=pulse-pro` for `?model=pulse`. See the [Pulse model card](/models/model-cards/speech-to-text/pulse) for the full language list.
## Step 3: Build It Into Your App
The snippets below use **Pulse Pro** (`?model=pulse-pro`). For multilingual audio, swap to `?model=pulse` and set the appropriate `language` code. Both models share the same `/waves/v1/stt/` endpoint and request shape.
```bash cURL
curl -sL "https://github.com/smallest-inc/cookbook/raw/main/speech-to-text/getting-started/samples/audio.wav" | \
curl -X POST "https://api.smallest.ai/waves/v1/stt/?model=pulse-pro&language=en" \
-H "Authorization: Bearer $SMALLEST_API_KEY" \
-H "Content-Type: application/octet-stream" \
--data-binary @-
```
```python Python
import os
import requests
API_KEY = os.environ["SMALLEST_API_KEY"]
SAMPLE_URL = "https://github.com/smallest-inc/cookbook/raw/main/speech-to-text/getting-started/samples/audio.wav"
audio = requests.get(SAMPLE_URL).content
response = requests.post(
"https://api.smallest.ai/waves/v1/stt/",
params={"model": "pulse-pro", "language": "en"},
headers={
"Authorization": f"Bearer {API_KEY}",
"Content-Type": "application/octet-stream",
},
data=audio,
timeout=120,
)
response.raise_for_status()
print(response.json()["transcription"])
```
```javascript JavaScript
const audioResponse = await fetch("https://github.com/smallest-inc/cookbook/raw/main/speech-to-text/getting-started/samples/audio.wav");
const audio = Buffer.from(await audioResponse.arrayBuffer());
const params = new URLSearchParams({ model: "pulse-pro", language: "en" });
const response = await fetch(`https://api.smallest.ai/waves/v1/stt/?${params}`, {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.SMALLEST_API_KEY}`,
"Content-Type": "application/octet-stream",
},
body: audio,
});
if (!response.ok) throw new Error(await response.text());
console.log((await response.json()).transcription);
```
For Pulse Pro, `language` must be `en`. For Pulse, set `language` to the known code (`en`, `hi`, `es`, etc.) for best accuracy, or use a regional aggregator for unknown audio. Aggregators differ by mode: pre-recorded supports `multi-eu` and `multi-asian`; streaming supports `north_indic`, `multi-asian` (US region only, contact sales for India), and `multi-south-indic` (India region only, Beta). See the [Pulse model card](/models/model-cards/speech-to-text/pulse) for the full per-mode language matrix.
**Full runnable source files:** [Python](https://github.com/smallest-inc/cookbook/blob/main/speech-to-text/transcribe-python.py) | [JavaScript](https://github.com/smallest-inc/cookbook/blob/main/speech-to-text/transcribe-javascript.js) | [cURL](https://github.com/smallest-inc/cookbook/blob/main/speech-to-text/transcribe-curl.sh)
## Step 4: Explore Features
#### [Real-Time Transcription](/models/documentation/speech-to-text-pulse/realtime-web-socket/quickstart)
Stream audio via WebSocket for live transcription.
#### [Speaker Diarization](/models/documentation/speech-to-text-pulse/features/diarization)
Identify and label different speakers.
#### [Word Timestamps](/models/documentation/speech-to-text-pulse/features/word-timestamps)
Precise timing for each transcribed word.
#### [Emotion Detection](/models/documentation/speech-to-text-pulse/features/emotion-detection)
Analyze emotional tone in speech.
#### [Model card](/models/model-cards/speech-to-text/pulse)
Benchmarks, capabilities, pricing.
Full endpoint spec: [Pulse API Reference](/waves/api-reference)
## Need Help?
#### [Join Discord](https://discord.gg/9WtSXv26WE)
Ask questions and get help from the community.
Or email [support@smallest.ai](mailto:support@smallest.ai).