Free Speech-to-Text APIs let developers add voice transcription to apps without building models from scratch. Palabra extends this with real-time translation alongside transcription.
Free Speech-to-Text APIs Overview
What is Speech-to-Text API Integration?
STT APIs convert audio streams or files into text using cloud-based AI models. Developers send audio via HTTP/WebSocket, receive JSON with timestamps, speaker labels, and confidence scores. Palabra’s STT works the same way but adds live translation to 60+ languages simultaneously.
Why Integrate STT with Palabra?
•Multilingual by default — transcribe English, translate to Spanish/French live
•Real-time streaming — <500ms latency for live captions
•Zoom/MS Teams and Google Meet native — no separate audio routing needed
•Developer-first — Python/JS SDKs, REST + WebSocket endpoints
Palabra’s STT Integration Options
Palabra Native STT API
Palabra’s built-in STT handles transcription + translation in one API call:
POST https://api.palabra.ai/v1/transcribe
{
“audio”: “base64…”,
“source_lang”: “en”,
“target_languages”: [“es”, “fr”, “de”]
}
7-day free trial available at app.palabra.ai. Paid plans start at Pro (150 credits/month). Credits are charged per minute of usage; rates vary by product type and plan tier.
Compatible 3rd Party Free APIs
•OpenAI Whisper — best accuracy, no real-time
•Google Speech-to-Text — 60 free min/month
•AssemblyAI — 100 free min/month
•Deepgram — 200 free min/month
Integration Quickstart
1-Minute Palabra STT Setup
1.Sign up: app.palabra.ai — get your API key
2.Install SDK: pip install palabra-client
3.Test call:
import palabra
client = palabra.Client(“your_key”)
result = client.transcribe_file(
audio=”meeting.mp3″,
target_languages=[“es”, “fr”]
)
Code Examples (Python/JS)
Real-time streaming (Python):
stream = client.stream_transcribe(
microphone=True,
target_languages=[“es”]
)
for chunk in stream:
print(chunk[“text_es”])
Browser integration (JS):
const stream = await palabra.transcribeStream(navigator.mediaDevices);
Performance Comparison
| API | Real-Time | Multilingual | Latency | Free Access |
| Palabra | Yes | Yes | <500ms | 7-day trial |
| Google STT | Yes | Yes | 1-2s | 60 min/month |
| Whisper | No | Yes | N/A | 10 min/month |
| Deepgram | Yes | No | 300ms | 200 min/month |
| AssemblyAI | Yes | Partial | 800ms | 100 min/month |
Step-by-Step Integration Guide
Connect to Existing STT
4.Configure Palabra pipeline: Set stt_provider: “custom”
5.Pipe audio: Palabra → your STT → back to translation
6.Merge outputs: Single JSON with transcription + translations
Build Custom STT App
Zoom -> Palabra STT -> Language Detection -> Translation -> Captions
Trial & Upgrade Path
What You Get on Trial (7 Days)
•Full platform access during the 7-day free trial
•All 60+ languages + regional variants
•Full REST/WebSocket API access
•Python/JS SDKs
•Zoom/Teams/Webex integrations
When to Upgrade
•More than 100 min/month usage
•Custom glossaries (product names)
•<200ms priority latency
•SOC2/enterprise security
•Unlimited storage
Use Cases & Code Samples
Live meeting transcription:
result = client.transcribe_stream(
audio_source=”zoom://meeting-123″,
target_languages=[“es”, “fr”, “de”]
)
Voice agent with captions:
palabra.transcribe(navigator.mediaDevices)
.then(captions => displayMultilingual(captions));
