AI Model Wiki

    Skip to main content
    7 min read
    Lesson 2 of 5

    ElevenLabs: Deep Dive

    From the course AI Voice Cloning & Text-to-Speech

    ElevenLabs has set the standard for AI voice quality. Let's explore its features in depth.

    Getting Started

    Sign up at elevenlabs.io. Free tier includes 10,000 characters per month (~10 minutes of audio).

    Pricing tiers:

    • Free: 10K characters/month, 3 custom voices
    • Starter ($5/month): 30K characters, 10 custom voices
    • Creator ($22/month): 100K characters, 30 custom voices, commercial license
    • Pro ($99/month): 500K characters, 160 custom voices, highest quality

    Text-to-Speech Basics

    1. Navigate to the Speech Synthesis page
    2. Select a voice from the library (or create your own)
    3. Paste or type your text
    4. Adjust settings:
      • Stability: Lower = more expressive, Higher = more consistent
      • Clarity + Similarity Enhancement: Higher = closer to original voice
      • Style Exaggeration: Amplifies the voice's emotional range
    5. Click "Generate" and listen
    6. Download as MP3 or WAV

    The Voice Library

    ElevenLabs offers thousands of pre-made voices:

    • Filter by gender, age, accent, and use case
    • Preview voices before selecting
    • Community-created voices available
    • Categories: Narration, Characters, News, Conversational, etc.

    Voice Settings for Different Content

    Audiobook narration:

    • Stability: 60-75% (allows natural variation)
    • Clarity: 75-85%
    • Style: 20-40% (subtle expressiveness)

    Marketing voiceover:

    • Stability: 70-80%
    • Clarity: 85-95%
    • Style: 30-50% (energetic)

    Character dialogue (games/animation):

    • Stability: 30-50% (highly expressive)
    • Clarity: 70-80%
    • Style: 50-80% (dramatic range)

    Projects: Long-Form Audio

    For audiobooks or long content, use the Projects feature:

    1. Create a new Project
    2. Upload or paste the full text
    3. Assign different voices to different characters/sections
    4. Generate chapter by chapter
    5. Fine-tune individual paragraphs
    6. Export the complete audio

    Projects handle text splitting, pacing, and consistency automatically.

    Speech-to-Speech

    Record yourself speaking and ElevenLabs transforms it into the selected AI voice while preserving your natural intonation, pacing, and emotion. This produces the most natural-sounding results because human prosody guides the AI.

    Sound Effects Generation

    ElevenLabs also generates sound effects from text descriptions:

    • "A thunderstorm with heavy rain"
    • "Footsteps on gravel"
    • "A crowd cheering in a stadium"

    Key Takeaways

    • •Free tier offers 10K characters/month — enough to experiment extensively
    • •Stability, Clarity, and Style settings dramatically affect output quality
    • •Use Projects for long-form audio like audiobooks with multi-voice support
    • •Speech-to-Speech preserves human intonation for the most natural results
    • •Different content types (narration, marketing, dialogue) need different voice settings
    0 of 5 lessons marked complete

    Progress is kept in this browser only.