Native, synchronized sound
With audio on, footsteps, ambience, music and spoken lines are generated with the picture and line up with the action. Switch audio off when you plan to add your own soundtrack in an editor.
ByteDanceHot
ByteDance's flagship video model with native audio
Seedance 2.0 generates cinematic 4–15 second clips with synchronized sound, strong motion and camera control. Use start/end frames or up to 9 reference images for consistent characters and products.
Try Seedance 2.0Hot prompts from the community — open any of them and remix it in one click.
FAQ
It starts at 80 credits; the default setup is 198 credits. Resolution and duration change the price, and you see the exact cost before generating. Failed generations are refunded automatically. New accounts get 50 free credits, and Pro includes 1,000 credits per month for $9.90.
Yes. Videos you generate can be used in ads, online stores and social media, subject to NeoMundo's terms of service and the model provider's policies.
Frames mode uses your image as the first frame, or two images as first and last frames, so the shot starts exactly on your picture. Reference mode treats up to 9 images as a guide to who or what appears, and the model builds a new composition around them.
Yes, when audio is switched on. Sound effects, ambience, music and dialogue are generated together with the picture. You cannot upload your own audio file to this model.
Draft at 480p or 720p with 4 or 5 seconds to check motion and framing cheaply. Once the prompt works, render the final at 1080p with the duration you need, up to 15 seconds.
Use the Fast variant for high-volume drafts and social clips at 480p or 720p. Use the full model when you need 1080p output for ads, client work or a final edit.
Seedance 2.0 offers the widest range of durations (4 to 15 seconds), 7 aspect ratios and up to 9 images. Veo 3.1 goes up to 4K but stops at 8 seconds; Kling 3.0 offers 5, 10 or 15 seconds.
Write one shot with a camera move and a sound, add a photo if you have one, and watch it play in a few minutes. Your free sign-up credits cover a first draft.
With audio on, footsteps, ambience, music and spoken lines are generated with the picture and line up with the action. Switch audio off when you plan to add your own soundtrack in an editor.
Ask for a dolly-in, an orbit, a handheld follow or a crane up and the camera does it, while bodies, fabric and liquids move with believable weight. This is the model to pick for action, dance and product moves.
Frames mode pins your image as the opening shot, or uses two images as start and end frames. Reference mode takes up to 9 reference images of a character, product or location and keeps them consistent inside a new scene.
Choose 4, 5, 6, 8, 10, 12 or 15 seconds. Five seconds suits a product reveal, ten fits a spoken UGC review, and fifteen gives room for a short scene with a beginning, middle and end.
Render at 480p, 720p or 1080p in 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 or adaptive, which follows the shape of your uploaded frame so nothing gets cropped.
Seedance 2.0 is the flagship video generation model from ByteDance. It creates cinematic clips of 4 to 15 seconds with synchronized sound, strong motion and precise camera control. On NeoMundo it sits in the video studio next to Veo 3.1, Kling 3.0 and Hailuo H3, and every model draws from the same credit balance.
A text prompt is required, and you can add up to 9 images in JPG, PNG or WebP, each up to 10 MB. With no images it is text to video. In frames mode, one image becomes the first frame, which is classic image to video, and a second image sets the last frame. In reference mode, the images act as a cast list: the same face, outfit or product reappears in a scene the model composes. Video and audio uploads are not accepted; the soundtrack is generated.
Seedance 2.0 starts at 80 credits for the smallest setup, and the default configuration costs 198 credits. Price scales with resolution and duration, and the studio shows the exact number before you press generate. As a reference point, the Unboxing recipe runs 8 seconds at 720p with audio for 317 credits, and the UGC Selfie Review runs 10 seconds for 396 credits.
Videos usually take 1 to 5 minutes, and you can queue up to 8 generations at once. For rough drafts, Seedance 2.0 Fast offers the same controls at 480p or 720p for fewer credits; switch back to the full model when you need 1080p for the final cut.
Click the Try button at the top of this page, or go to /create/video and choose the model. Sign in first; new accounts receive 50 free credits, and Pro adds 1,000 credits every month.
Cover four things in order: subject, action, camera move and sound. For example: a barista pours latte art in slow motion, the camera pushes in to the cup, soft café chatter and a clink of porcelain. Put spoken lines in quotes and say who speaks them.
Use frames mode to start from your exact photo or to connect a start and an end frame. Use reference mode with up to 9 reference images when a character or product must stay recognizable in a new setting.
Pick aspect ratio, resolution, duration and audio on or off. Test at 480p or 720p, then re-run your best prompt at 1080p and download the clip.
Upload a product photo and write a short script. The UGC Selfie Review and Unboxing recipes use this model to make a creator talk about your product or open the box on camera in vertical 9:16.
The Fashion Walk recipe shows a model walking and turning to reveal fit and drape; Product In Use places your item in a real-life scene. Both cost 198 credits at 720p for 5 seconds.
Fifteen seconds with sound is enough for a mini story: an establishing shot, an action beat and a reaction. Use 21:9 for a film look or 16:9 for YouTube.
Keep the same reference images for every episode of a series so faces and outfits stay stable from clip to clip, then join the clips in your editor.