Cartesia

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

We've raised a $64M Series A led by @kleinerperkins to build the platform for real-time voice AI. We'll use this funding to expand our team, and to build the next generation of models, infrastructu…

AI AgentsLaunchMarch 11, 2025@cartesia
0:00 · The Hook · Introducing Sonic 2: The Next Generation of Voice AI
0:00 / 0:00

Scene-by-scene timeline & spoken transcript

  1. The Hook

    Introducing Sonic 2: The Next Generation of Voice AI

    “We've been building the platform for real-time voice AI. We've raised a $64M Series A led by Kleiner Perkins to build the platform for real-time voice AI. We'll use this funding to expand our team, and to build the next generation of models, infrastructure, and tools for real-time voice AI. We're excited to announce Sonic 2, our next generation voice AI model. Sonic 2 is the fastest, most controllable, and most accurate voice AI model in the world.”

    On screen
    Karan Goel Founder / CEO @ Cartesia Update Olivia Moore $64M Posted January 29, 2025 AI YOU'RE THE VOICE YOU'RE THE VOICE AI Voice 2025: The Year Voice AI Finds Its Voice What makes voice AI all the rage now? It simplifies complexity. Getting things done easier than ever, all thanks to the power of voice AI. Why Voice AI Is The New U What makes voice AI all the rage now? It simplifies complexity. Getting things done is easier than ever, all thanks to the power of voice AI. Published on January 13, 2025 2024 te of Voice AI
    Camera
    Medium close-up on the founder, with occasional cuts to animated text and graphics. Camera remains static on the founder.
    Motion
    Text reveals with subtle scaling and opacity changes, combined with quick cuts to founder speaking. Data visualizations animate in with wipe effects.
  2. Product Reveal

    Unmatched Speed and Accuracy

    “Sonic 2 is the fastest model in the world. We're talking about 40 milliseconds of latency. That's faster than any other model out there. And it's not just fast, it's also incredibly accurate. We've achieved best-in-class voice cloning, and we're able to handle complex phrases and accents with ease.”

    On screen
    Win Rate Blinded Human Evals Cartesia 61.4% ElevenLabs 38.6% Cartesia 74.2% Blinded, head-to-head human preference evaluations on 100 held out voices. Voice cloning not supported on OpenAI, Azure, Google T Model Latency Sonic 1 Sonic 1.0 135 ms Sonic 2.0 90 ms Sonic 2.0 Turbo 40 ms P90 Latency (ms) 90th percentile latency (including all network and serving latency) from different datacenter locations. Sonic 2.0 ElevenLabs Flash V2 PlayHT 3.0-mini OpenAI Asia Pacific Sonic 2.0 ElevenLabs Flash V2 PlayHT 3.0-mini OpenAI United States Sonic 2.0 ElevenLabs Flash V2 PlayHT 3.0-mini OpenAI Europe
    Camera
    Alternates between founder close-up and full-screen data visualizations. Camera on founder is static.
    Motion
    Bar charts and pie charts animate in with smooth transitions. Text highlights key data points. Pixel art character animates with waveform.
  3. Feature Teaser

    Advanced Control and Native Integration

    “Spelling is crucial for voice agents. We've built in advanced control features, allowing you to dictate how words are pronounced and spelled. And Sonic 2 runs natively on device, meaning you can integrate it directly into your applications for unparalleled performance.”

    On screen
    Spelling is crucial for voice agents names e-mails phone numb Your confirmation number is AT121 best in class Voice cloning Hey Jonathan, you should check out Sonic 2.0. Sonic 2.0 runs natively
    Camera
    Mix of founder close-up and animated feature demonstrations. Founder camera is static.
    Motion
    Text-based feature explanations with iconographic support. Pixel art characters demonstrate voice agent interaction. Text reveals with a subtle glow effect.
  4. Call to Action

    Experience the Future of Voice AI

    “(No spoken dialogue — The audio features a consistent electronic score with a driving beat. It incorporates pizzicato synths, sub-bass, and various sound design elements to enhance transitions and emphasize key points. There are high-energy audio transitions and risers at multiple points, culminating in impactful sub-bass drops. The overall soundscape is modern and dynamic, supporting the product launch's energetic tone.)”

    On screen
    (No on-screen text)
    Camera
    Dynamic cuts between various abstract UI elements and product shots. Camera motion is fluid and often involves subtle zooms or pans.
    Motion
    Abstract UI elements animate with glowing lines and particles. Product shots feature smooth camera movements and focus shifts. High-energy transitions with sound design.

Related AI Agents Product Launches

Explore all AI Agents launches →
OpenAI
Hook 9.2137.9M
OpenAIAI Agents

To ensure that artificial general intelligence benefits all of humanity

@OpenAI
Gopuff
Hook 9.277.6M
GopuffAI Agents

Gopuff introduces Go, an AI shopping assistant built with SpaceXAI: say what you need and the order is placed.

@gopuff
Elon Musk
Hook 9.263.3M
Elon MuskAI Agents

Iliad (Troy) trailer made by Grok Imagine 1.5, which was just released

@elonmusk