Pre-Recorded Audio

Upload pre-recorded audio files for high-quality batch transcription. Perfect for podcasts, meetings, interviews, and media content.

Batch Endpoints

POST/api/v1/stt/transcribe

Universal endpoint - automatically routes to the best engine based on selected model.

POST/api/v1/stt/cipher/transcribe

Cipher engine endpoint - use for cipher-fast and cipher-max models.

POST/api/v1/stt/lucid/transcribe

Lucid engine endpoint - use for lucid-mono, lucid-multi, lucid-agent, lucid-lite models.

Cipher Models

Optimized for batch transcription with excellent accuracy and speed.

ModelDescriptionCredits/secBest For
cipher-fastUltra-fast transcription2Quick turnaround, general content
cipher-maxMaximum accuracy2Complex audio, accents, technical content

Cipher Features

  • Language Detection: Automatic detection or specify from 90+ languages
  • Word Timestamps: Granular word-level timing for precise alignment
  • Context Prompt: Guide the model with domain-specific terms and phrases
  • Multiple Formats: JSON, text, verbose_json, SRT, VTT output formats

Lucid Models (Batch)

Enterprise-grade batch transcription with advanced features.

ModelDescriptionCredits/secBest For
lucid-monoHigh-accuracy single language3Single-language content
lucid-multiMultilingual support3Mixed-language content
lucid-agentConversational optimized3Phone calls, support recordings
lucid-liteCost-effective option3High volume processing

Lucid Features

  • Speaker Diarization: Identify and label different speakers in conversations
  • Keywords Boost: Improve recognition accuracy for specific words
  • Smart Formatting: Auto-format dates, numbers, and currency
  • Punctuation: Automatic punctuation insertion

Basic Request Example

import requests

url = "https://yourvoic.com/api/v1/stt/transcribe"
headers = {"X-API-Key": "your_api_key"}

with open("audio.mp3", "rb") as f:
    response = requests.post(url, headers=headers,
        files={"file": f},
        data={"model": "cipher-fast"})

result = response.json()
print(result["text"])