ElevenLabs

What is ElevenLabs?

Company and Funding

• 2023: $4.6M - $25M
• 2024: $80M - $120M
• 2025: $200M+ ARR (August 2025)

• Pre-Seed (January 2023): $2M (Credo Ventures)
• Series A (June 2023): $19M ($100M val.)
• Series B (January 2024): $80M ($1.1B val. - Unicorn)
• Series C (January 2025): $180M ($3.3B val.)
• Secondary (September 2025): $100M tender ($6.6B val.)

Available Voice Models (December 2025)

Eleven v3 (June 2025) - Most Advanced

• ElevenLabs' most expressive model
• 70+ languages supported
• Inline audio tags: [excited], [whispers], [sighs], [laughs]
• Text to Dialogue: natural multi-speaker conversations
• Automatic interruptions and tone shifts
• Limit: 40,000 characters/call
• 1 credit = 1 character
• Ideal for: audiobooks, cinematic content, media
• NOT recommended for real-time (high latency)

Eleven Multilingual v2

• Most realistic and emotionally rich voices
• 29 languages supported
• Ideal for voiceovers, audiobooks, content
• Limit: 10,000 characters/call
• 1 credit = 1 character
• Premium quality, moderate latency

Eleven Flash v2.5 - Ultra-Fast

• Ultra-low latency: ~75ms
• 32 languages (adds Vietnamese, Hungarian, Norwegian)
• 3x faster for non-English languages
• 25% faster for English
• 0.5 credits = 1 character
• Ideal for: conversational agents, chatbots, gaming, real-time

Eleven Flash v2

• English only
• ~75ms latency
• 0.5 credits = 1 character
• Maximum speed for English use cases

Pricing and Plans (December 2025)

Credit System

• Eleven v3: 1 credit = 1 character
• Multilingual v2: 1 credit = 1 character
• Flash v2/v2.5: 0.5 credits = 1 character
• ~1,000 credits ≈ 1 minute of audio

Plan Details

• 10,000 credits/month (~10 min)
• Basic TTS in 32 languages
• Simple voice cloning
• Basic API access
• No commercial license
• Attribution required

• 30,000 credits/month
• Commercial license
• Basic voice cloning
• Basic Dubbing Studio
• Email support

• 100,000 credits/month (~2 hours)
• Professional Voice Cloning
• Full Dubbing Studio
• Projects for long-form content
• Full API access
• No watermark

• 500,000 credits/month (~10 hours)
• Higher quality output
• Increased concurrency
• Priority support
• Enterprise use

• 2,000,000 credits/month (~40 hours)
• Maximum concurrency
• Dedicated support
• For teams and agencies

• Custom credits
• SSO/SAML
• HIPAA compliance (add-on $1,000/mo)
• Dedicated SLA
• Custom integrations

API Pricing

• ~$0.20 per 1,000 characters (standard voices)
• Volume pricing available
• Concurrency: 4-30 sessions depending on plan

Products and Features

Text to Speech (TTS)

• Most realistic voices on the market
• Natural emotions and intonation
• Stability and similarity control
• Audio tags for fine control
• Multiple output formats

Voice Cloning

• Instant Voice Clone (IVC): 1 minute of audio
• Professional Voice Clone (PVC): higher fidelity
• Unique characteristics preservation
• Supported in v3, v2, Flash

AI Dubbing Studio (2024)

• Automatic translation to 32 languages
• Preserves original voice, emotion and timing
• Lip synchronization
• Speaker differentiation
• Noise removal included

Voice Design

• Generate custom voices from text description
• No reference audio needed
• Voice characteristics control

Conversational AI (November 2024)

• Platform for interactive voice agents
• Real-time latency
• 2M+ agents created
• Customer support, gaming, education

Sound Effects Model

• AI-generated sound effects
• Expansion beyond voice

Voice Isolator (July 2024)

• Removes background noise
• Improves audio quality

ElevenReader App (June 2024)

• iOS and Android
• Converts articles, PDFs, ePubs to audio
• High-quality AI voices
• Audiobook publishing (February 2025)

Eleven Music (August 2025)

• Full music generation from text
• With or without vocals
• Multiple languages (English, Spanish, German, Japanese)
• Pre-cleared for commercial use
• Genre, style, structure control
• Section-based editing

Voice Library & Marketplace

• 5,000+ voices shared by creators
• Voice actors earned $2M+ in rewards
• Iconic Voices: Michael Caine, Matthew McConaughey (November 2025)

Scribe (February 2025)

• Speech-to-text model
• Character-level timestamps
• Speaker diarization
• Industry-leading word error rate

AI Speech Classifier (June 2023)

• Detects if audio was generated by ElevenLabs
• API available
• Digital signature in generated audio

Notable Customers

Impact Program

• 80+ organizations in accessibility, education, culture
• Bridging Voice, The Scott Morgan Foundation (ALS)
• 1,000+ people with speech impairments reclaimed their voice
• KADIST and Centre Pompidou (Paris)
• Expansion in Poland and India

History and Achievements

• 2022: Founded by Dąbkowski and Staniszewski
• January 2023: Public beta, viral "Harry Potter by Balenciaga"
• June 2023: AI Speech Classifier, Series A
• August 2023: Expansion to 28 languages
• September 2023: Projects for long-form content
• October 2023: AI Dubbing to 20+ languages
• January 2024: Unicorn ($1.1B), Voice Marketplace
• June 2024: ElevenReader App
• July 2024: Voice Isolator
• November 2024: Conversational AI
• January 2025: Series C ($3.3B)
• February 2025: Scribe STT, Audiobook publishing
• June 2025: Eleven v3 (70+ languages, audio tags)
• August 2025: Eleven Music, $200M ARR
• September 2025: $6.6B valuation
• November 2025: Iconic Voices Marketplace

Controversies

• 4chan abuse (2023): Users generated hate messages with celebrity voices
• Biden Robocalls (January 2024): Fake AI calls in New Hampshire primary, attorney general investigation
• Deepfake concerns: Criticism about impersonation potential

Competition

Key Advantages

• Most realistic voices on the market
• Superior emotional range
• 70+ languages in v3
• High-fidelity voice cloning
• Developer-friendly API
• Complete ecosystem (TTS, STT, Music, Dubbing)
• <75ms latency (Flash)
• Monetized voice marketplace

Limitations

• Can be expensive for high volume
• v3 requires prompt engineering
• v3 not suitable for real-time
• PVC not optimized for v3 yet
• Complex credit system
• HIPAA is expensive add-on ($1,000/mo)