ElevenLabs

ElevenLabs · Launch Video Breakdown: Hook, Pacing & Motion Design

Today we’re introducing Scribe v2: the most accurate transcription model ever released. While Scribe v2 Realtime is optimized for ultra low latency and agents use cases, Scribe v2 is built for batch

0:00 · The Hook · Scribe v2 Announcement
0:00 / 0:00
AI AgentsPublic LaunchJanuary 9, 2026

Benchmark Breakthrough

9.6

Cuts / min

0:50

Runtime

558K

Views

0

Likes

251

Reposts

93

Replies

Scene-by-scene timeline & spoken transcript

  1. The Hook

    Scribe v2 Announcement

    “ElevenLabs speech-to-text just had another major breakthrough with Scribe v2.”

    On screen
    V2
    Camera
    Static centered composition with subtle grid morph
    Motion
    Grid tile expansion revealing an iridescent fluid sphere collapsing into a badge
  2. Product Reveal

    Benchmark Outperformance & Multilingual

    “Outperforming major ASR competitors across English and multilingual datasets, including dynamic detection for over 90 different languages.”

    On screen
    Gemini 2.5 Pro 93.9% GPT 4o Transcribe 94.1% ElevenLabs Scribe v2 95.7% Dynamic multilingual detection across 90+ languages.
    Camera
    Smooth vertical pan down from light to dark canvas
    Motion
    Linear data progress bars with bouncing orb triggers and cycling multilingual typography
  3. Feature Teaser

    Timestamps & Diarization

    “With improved character-level timestamps, speaker diarization...”

    On screen
    0:54 0:56 0:58 1:00 1:02 1:04 1:06 T4 T5 Detecting Voices Speaker 01 Speaker 02 Speaker 03
    Camera
    Static focus with scrolling timeline scrubber
    Motion
    Ruler tick scale alignment, line-by-line staggered transcript reveal with speaker avatar dots
  4. Feature Teaser

    Custom Keyterm Prompting

    “And now with custom keyterm prompting to guarantee consistent improvement across names, technical vocabulary, and domain specific terms.”

    On screen
    Custom Keyterm Prompting Mariusz Dąbrowski Rec.709 UHD LoRA
    Camera
    Horizontal kinetic drift tracking words
    Motion
    Floating pill dispersion, strikethrough error correction morphing into validated terms
  5. Feature Teaser

    Acoustic Accuracy

    “Word by word, syllable by syllable, captured with unmatched accuracy.”

    On screen
    Word by word Syllable by syllable Unmatched Accuracy
    Camera
    Static centered perspective
    Motion
    Kinetic text scaling accompanied by propagating concentric sonar wave rings
  6. Feature Teaser

    Entity Detection

    “And now with entity detection too, for the ability to collect and identify specific information with exact timestamps.”

    On screen
    Entity Detection Label Description account_number bank_account date_interval email_address driver_license credit_card 587 4988 2404 4680 3397 12/04/29
    Camera
    Vertical scrolling list transitioning to horizontal timeline pan
    Motion
    Slot-machine vertical selector wheel highlighted row, followed by inline text pill extraction
  7. Call to Action

    Outro & Website

    “So start building with Scribe v2 today. Only with ElevenLabs.”

    On screen
    Scribe v2 ElevenLabs elevenlabs.io/speech-to-text
    Camera
    Static center hold
    Motion
    Typographic kerning expansion and logo lockup resolution

Related AI Agents Product Launches

Explore all AI Agents launches →
OpenAI
@OpenAI
AI Agents
@OpenAI
OpenAI@OpenAI
Gopuff
@gopuff
AI Agents
@gopuff
Gopuff@gopuff
Elon Musk
@elonmusk
AI Agents
@elonmusk
Elon Musk@elonmusk