Breathing Life into Characters — WaveSpeedAI at Agora Convo AI World Japan
On November 5, 2025, we joined Agora’s “Convo AI World – Japan”, an event all about how conversational AI, anime, and avatars are coming together to create digital characters that can actually talk back. Hosted by Agora and V-Cube, it brought together developers, creators, and innovators who are shaping the next wave of interactive experiences.
Finding the Sweet Spot: Speed and Quality
During a fireside chat moderated by Patrick Ferriter, our founder Zeyi Cheng talked about one of the biggest challenges in real-time AI — finding the balance between performance and visual quality.
“We’re seeking a balance between speed and output quality,” Zeyi shared. “Our customers are asking for near real-time diffusion-based avatar models — and we’re working to make that happen.”
Patrick Ferriter added his perspective on this challenge, noting that:
“The need for real-time digital humans with both exceptional speed and high fidelity is the next major hurdle in conversational AI. WaveSpeedAI’s InfiniteTalk engine is addressing this challenge head-on, delivering the kind of innovation that will accelerate the adoption of these interactive experiences across the industry and within the broader developer ecosystem.”
That’s exactly what we’re building with Infinitetalk, our digital human engine. With just one image and a short voice clip, anyone can create a lifelike virtual human that talks, moves, and emotes — all in real time. For Japan’s vibrant anime and VTuber scene, this opens up a whole new world of creativity. No studio setup. No long production cycle. Just imagination and a few clicks.
Multimodal Ecosystem
At the event, we met a lot of creators excited about our audio generation and processing capabilities. We also shared our plan to help a local Japanese company publish its audio model on the WaveSpeedAI platform — a step toward building a more open and collaborative AI ecosystem.
Looking Ahead
As Zeyi put it:
“Soon, avatar videos will be indistinguishable from real people. Costs will fall, and adoption will spread globally.”
That’s the future we’re chasing — one where AI helps every creator bring their characters to life, instantly and beautifully.
Try WaveSpeedAI InfiniteTalk Today
Come experience WaveSpeedAI’s digital humans today! You’ll be amazed by the lifelike expressions, natural movements, and perfectly synced lip motions — it feels just like talking to a real person!
👉 infinitetalk 👉 multi-infinitetalk 👉 infinitetalk（video-to-video）
Stay Connected With Us
Discord Community | X (Twitter) | Open Source Projects | Instagram
© 2025 WaveSpeedAI. All rights reserved.
