ByteDance launches Seedance 1.5 Pro on Doubao, democratizing AI-Powered talking video creation

ByteDance has officially rolled out its next-generation audiovisual generation model—Seedance 1.5 Pro—on Doubao, its AI-powered assistant platform. The update empowers everyday users to create rich, synchronized talking videos from simple text prompts—no editing skills required.

At the heart of Seedance 1.5 Pro is an end-to-end multimodal pipeline that unifies text, visuals, and audio into a single generative process. Unlike older systems that generate video and voice separately—often leading to mismatched timing or tone—this model deeply interprets semantic intent to simultaneously produce coherent scenes, natural-sounding dialogue, and context-aware sound effects.

Key advancements include:

• True audio-visual synchronization: Dialogue, ambient sounds, and visuals are generated in harmony, eliminating “lip-sync drift” or emotional dissonance.
• Human-like speech delivery: Voices now feature nuanced intonation, pacing, and expressiveness that closely mimic real human narration.
• Cinematic storytelling: Dynamic camera movements, shot transitions (e.g., close-ups, pans), and complex character interactions—such as hand gestures or facial expressions—are rendered automatically, without manual post-production.

For users, the workflow is remarkably simple:

1. Open the Doubao app and tap the “Animate Photo” feature.
2. Select the “1.5 Pro” model.
3. Upload a reference image—whether a selfie, sketch, or illustration.
4. Enter a prompt like: “Have this cat tell a bedtime story in a gentle voice, set in a starry bedroom.”

Within seconds, a fully narrated, animated video is generated—complete with synchronized lip movements, expressive tone, and atmospheric visuals.

With Seedance 1.5 Pro, Doubao evolves from a text-and-image creator into a one-stop studio for dynamic storytelling. Whether crafting short narratives, product demos, or breathing life into static art, users can now turn ideas into immersive audiovisual content—directly within a chat interface.

This release marks another leap toward ByteDance’s vision: making high-quality video creation as effortless as typing a sentence.