Generate cinematic video up to 15 seconds long — 480p to 1080p, with 4K available — featuring multi-shot consistency, multi-modal reference inputs, camera direction and native audio generation.
See what creators are making with Seedance 2.0. From cinematic camera moves to fluid character animations.
Turn your ideas into stunning AI videos using ByteDance Seedance 2.0 in just three simple steps.
To get started, select the Seedance 2.0 AI video model as your starting point.
Enter a detailed prompt or upload a reference image, video, or audio file to define the style, mood, and visuals of your project. You can upload up to 12 files across different formats to guide the output.
Review the generated results and adjust your prompt and style to align with your vision.
Seedance 2.0 on HeyOz lets you combine text, images, video, and audio to create cinematic AI videos with more creative control than prompting alone. It holds faces, scenes, styles, and motion steady across your whole video. You can edit or extend clips and generate synchronized audio for professional results.
Seedance 2.0 on HeyOz's AI video generator combines text, images, videos, and audio to guide your video across multiple scenes. Upload up to 12 reference files — as many as 9 images, 3 videos, and 3 audio clips — then point to each one by name in your prompt with @Image1 or @Audio1. Your actual product, your actual voiceover, exactly where you want them.
You can use Seedance 2.0 for full control over lighting, shadows, and performance, while directing advanced camera moves like tracking, dolly, POV, and rack focus shots. Provide a reference video and Seedance 2.0 will draw on its composition, motion, and camera language to guide your generation.
Generate synchronized dialogue, sound effects, and music that match your scene — with lip-synced delivery driven by your own voiceover track, produced natively alongside your video. Music, ambience, and voiceover arrive as separate tracks, so you can rebalance the mix without regenerating the shot.
Characters and key objects hold their look across cuts, angle changes, and lighting shifts — in a single generation rather than a stitched sequence. Need one element changed? Edit that clip and extend it, without re-rolling a spot you already approved.
From product ads built on your real photos to multi-shot brand films, in every format your channels need.
Seedance 2.0 is ByteDance's multimodal video model, first launched in China in February 2026 and available through the API since April 2026. It generates video and synchronized audio together in a single pass, and it's the only major model that accepts text, images, video, and audio all as reference inputs for one generation. On HeyOz it's wired straight into your brand kit and ad formats.
Pick Seedance 2.0 as your model, write a prompt or drop in your reference files, then generate and adjust. If you upload references, you can point to them individually in the prompt — @Image1 for your product shot, @Audio1 for your voiceover.
That's what HeyOz is built for. Point it at your product photos and the spot comes back on-brand and sized for Meta, TikTok, and more.
You can start creating on HeyOz for free. See our pricing page for what's included on each plan.
Between 4 and 15 seconds per generation, or set it to auto and let the model decide. The whole clip can carry multiple shots — cuts, angle changes, and scene transitions all happen inside that single generation, so there's no stitching.
Text, plus up to 12 reference files in total — as many as 9 images, 3 videos, and 3 audio clips. Video references can total 2 to 15 seconds and audio up to 15 seconds, and if you supply audio you'll need at least one image or video reference alongside it. You can also edit or extend a clip you've already generated.
480p, 720p, 1080p, and 4K. ByteDance's technical paper describes 480p and 720p as the model's native output resolutions, with the higher tiers available on top — so if you're finishing an ad for paid placement, generate at the tier you intend to ship rather than upscaling later.
Three variants, same family. Seedance 2.0 is the full model for final quality, Fast trades a little fidelity for quicker turnarounds, and Mini is the cheap option for volume. Draft your angles on Mini or Fast, then finish on the full model.
Pick the right engine for the job — all in one place.
Anyone can give you Seedance 2.0. Only HeyOz turns it into your ad.
HeyOz reads your brand and keeps every asset on-color, on-voice, on-message.
Seedance, Veo, Kling, GPT Image, and more — no juggling subscriptions.
Sized and formatted for Meta and TikTok, straight out of the box.
Not just a model — templates, avatars, and an agent that does it all.
Start Now. No agency, no brief, no blank screen.