Skip to content

MurfAI.md

Latest commit

 

History

History
230 lines (185 loc) · 7.99 KB

MurfAI.md

File metadata and controls

230 lines (185 loc) · 7.99 KB

Murf AI TTS Provider Setup Guide

Overview

Murf AI is a leading AI voice synthesis platform offering ultra-fast, high-quality text-to-speech with 150+ voices across 35+ languages. Features include Gen2 for content generation and Falcon for real-time voice agents.

Features

  • Premium Voice Quality: State-of-the-art Gen2 and Falcon models with natural-sounding prosody
  • Multilingual Support: 35+ languages with MultiNative technology for authentic multilingual content
  • Extensive Voice Library: 150+ voices across genders, ages, and accents
  • Multiple Models: Gen2 for studio-quality content, Falcon for ultra-low latency (<130ms)
  • Speaking Styles: 20+ styles including Conversational, Promo, Newscast, Storytelling, Calm, and more
  • Advanced Customisation: Control pitch, speed, emphasis, pauses, and pronunciation

Prerequisites

  • Murf AI account (free trial available)
  • Valid API key from Murf AI API Dashboard

Setup Instructions

1. Create Murf AI Account

  1. Visit https://murf.ai/
  2. Click Sign Up in the top-right corner
  3. Create account using:
    • Email and password
    • Google authentication
    • Other authentication methods
  4. Verify your email address if prompted

2. Get Your API Key

  1. Log in to your Murf AI account
  2. Navigate to https://murf.ai/api/dashboard
  3. Click Generate API Key
  4. Copy the generated API key
  5. Important: Store your API key securely - treat it as a password

3. Configure in TextToSpeech-Generator

  1. Launch the TextToSpeech-Generator application
  2. Select Murf AI from the TTS Provider dropdown
  3. Click the Configure button
  4. Paste your API key in the API Key field
  5. Click Test Connection to verify credentials
  6. Click Save & Close to complete setup

Pricing & Limits

Free Trial

  • Available for new accounts
  • Limited character quota for testing
  • Access to all voices and features

Paid Plans

  • Starter: For individuals and small teams
  • Pro: For content creators and businesses
  • Enterprise: Custom pricing for high-volume use

Visit https://murf.ai/pricing for current pricing.

Available Voices

Voice Samples

Murf AI offers 150+ voices including:

  • Natalie (Female, US): Natural, warm voice for general content
  • Ken (Male, US): Professional, authoritative voice
  • Terrell (Male, US): Deep, engaging voice
  • Julia (Female, US): Clear, conversational voice
  • Matthew (Male, UK): British accent, professional tone
  • And many more across 35+ languages

Voice Models

  • Gen2: Advanced model for studio-quality content generation, trained on 70,000+ hours of audio
  • Falcon: Ultra-fast model for real-time voice agents with <130ms latency

Speaking Styles

  • Conversational
  • Promo
  • Newscast
  • Storytelling
  • Calm
  • Furious
  • Angry
  • Sobbing
  • Sad
  • And more (voice-specific)

Configuration Options

Basic Settings

  • API Key: Your Murf AI API key (required)
  • Voice: Select from 150+ available voices
  • Language/Locale: 35+ supported languages and regional variants
  • Format: WAV, MP3, FLAC, PCM, OGG

Advanced Settings

  • Model: Gen2 (default) or Falcon for ultra-low latency
  • Speaking Style: Choose from 20+ voice-specific styles
  • Pitch: Adjust voice pitch (-50 to +50)
  • Speed/Rate: Control speaking speed (-50 to +50)
  • Sample Rate: 8000, 24000, 44100, or 48000 Hz
  • Channel Type: MONO or STEREO
  • Variation: Add natural variation in pause, pitch, and speed (0-5)

MultiNative Feature

Enable voices to speak in multiple languages while maintaining authentic pronunciation patterns for each language.

Supported Output Formats

Format Description Use Case
WAV Uncompressed audio (default) Low-latency applications
MP3 Compressed, widely supported General use, file size matters
FLAC Lossless compression High fidelity, smaller than WAV
PCM Raw uncompressed audio Telephony, DSP pipelines
OGG Efficient compression Web playback, streaming
ALAW Telephony compression Phone systems (8kHz mono only)
ULAW Telephony compression Phone systems (8kHz mono only)

API Endpoints

Base URL

https://api.murf.ai

Main Endpoints

  • POST /v1/speech/generate - Generate speech (non-streaming)
  • POST /v1/speech/stream - Generate speech (streaming, real-time)
  • GET /v1/speech/voices - List available voices

Troubleshooting

Authentication Errors (401 Unauthorized)

  • Cause: Invalid or expired API key
  • Solution:
    1. Verify API key is correctly entered (no extra spaces)
    2. Generate a new API key from the API Dashboard
    3. Ensure your account is active and in good standing

Rate Limiting (429 Too Many Requests)

  • Cause: Exceeded API rate limits
  • Solution:
    1. Check your rate limits in the API Dashboard
    2. Implement exponential backoff in your requests
    3. Consider upgrading your plan for higher limits

Character Quota Exceeded (402 Payment Required)

  • Cause: Exceeded monthly character allowance
  • Solution:
    1. Check remaining characters in API Dashboard
    2. Wait for quota reset (monthly)
    3. Upgrade plan for more characters

Voice Not Found

  • Cause: Invalid voice ID or voice not available in your region
  • Solution:
    1. Use GET /v1/speech/voices to list available voices
    2. Verify voice ID format (e.g., "en-US-natalie")
    3. Check if voice supports your selected language/locale

Timeout Errors

  • Cause: Request taking too long (>30 seconds)
  • Solution:
    1. Reduce text length for single requests
    2. Use streaming API for long-form content
    3. Check network connectivity

Invalid Format Combination

  • Cause: Format/sample rate/channel type combination not supported
  • Solution:
    1. ALAW/ULAW only support 8000 Hz mono
    2. Use WAV/MP3 for flexibility
    3. Check API documentation for format-specific restrictions

Best Practices

  1. API Key Security

    • Never share your API key
    • Store securely using Windows Credential Manager
    • Rotate keys periodically
    • Use environment variables in production

  2. Voice Selection

    • Test voices using the Voice Explorer
    • Consider accent and tone for your audience
    • Match voice to content type (promo vs. conversational)

  3. Performance Optimisation

    • Use streaming API for real-time applications
    • Leverage Falcon model for voice agents (<130ms latency)
    • Cache generated audio when appropriate
    • Use appropriate sample rates (lower for voice, higher for music)

  4. Content Quality

    • Use pronunciation dictionary for brand-specific terms
    • Add pauses and emphasis for natural speech
    • Leverage speaking styles for context-appropriate delivery
    • Test with your target audience

  5. Error Handling

    • Implement retry logic with exponential backoff
    • Monitor character usage to avoid quota issues
    • Log errors for debugging
    • Handle network timeouts gracefully

Additional Resources

Limitations

  • Maximum text length per request: 10,000 characters (Gen2)
  • Audio files from generate API available for 72 hours
  • Regional availability may vary
  • Some voices support more languages than others
  • Speaking styles are voice-specific

Security & Compliance

  • Enterprise-grade security
  • GDPR compliant
  • SOC 2 Type II certified
  • Data residency options available
  • Zero retention mode (Base64 encoding)

Note: Murf AI regularly updates their voice library and features. Check the official documentation for the latest capabilities and pricing.