We built Seedance 2.5 — a browser-based multi-modal AI video workspace

Hey everyone! We just launched Seedance 2.5 AI Video Generator — a browser-based AI video creation platform.

We built it because existing AI video tools struggle to keep characters, products and motion consistent across shots. Creators kept telling us they wanted one place to combine reference images, video clips and audio with natural language prompts, and get longer, controllable clips that actually match what they imagined.

What Seedance 2.5 does:

• Combine up to 9 reference images, 3 video clips and 3 audio tracks with text prompts

• - Generate cinematic videos with consistent faces, style, motion and sound

• - Extend and merge existing videos, replicate choreography and camera moves from reference footage

• - Built-in contextual audio generation

• - Runs in the browser with a free daily tier, no installation

We'd love feedback from this community — what do you think of multi-modal reference control for AI video? Any use cases you'd like to see supported?