Video model · ByteDance Seed
Seedance 2.0 is the video model from ByteDance’s Seed team, launched in February 2026. On Deepnia it creates 4–15 second videos where voices, sound effects and music are generated with the picture, from text, images or reference videos. Three variants: Mini to test fast, Fast for balance, Standard for quality up to 1080p.
The video tool opens with Seedance 2.0 already selected. The exact cost shows before every creation; you sign up when you generate.
Mini, Fast and Standard match the three official models in the Seedance 2.0 series. According to ByteDance, Standard aims for the best quality, Fast balances cost and speed, and Mini offers the best cost-performance. The Standard variant goes up to 1080p.
Voices, sound effects and background music are generated with the picture. Put dialogue lines in quotation marks and describe the sounds at the end of the prompt.
ByteDance highlights how Seedance 2.0 handles complex motion and interactions between characters and objects while following physical laws.
In Reference mode, combine up to 7 media, including up to 3 videos, to guide the character, the place, the style or the camera movement.
In Keyframe mode, set the start frame and, if you like, the end frame: Seedance 2.0 animates the move from one to the other.
The “Try Seedance 2.0” button opens Deepnia’s video tool with the model already selected.
Mini to validate an idea, Standard for the final cut in 1080p.
Stay with text only, or switch to Reference mode (images and videos) or Keyframe mode (start and end frames).
Choose the length, resolution and aspect ratio, check the cost shown, then create.
Copy a prompt, open the tool and adapt it to your project.
A 10-second vertical clip at night on a rooftop: a guitarist plays under string lights, couples dance slowly around him, the camera gently circles the musician. Sound: acoustic guitar, laughter, soft city noise. No on-screen text.
An 8-second ad: the bottle of hibiscus juice from the reference image stands on a wooden counter, ice cubes fall in slow motion into a glass, the ruby-red drink pours and sparkles in the sunset light. Sound: clinking ice, pouring liquid, short electronic beat. No added text or logo.
From the start frame (a closed leather bag on a wooden table) to the end frame (the open bag showing its lining), an artisan’s hands gently open the flap, the camera moves closer to the hand stitching. Warm natural light. Sound: soft leather creak, workshop ambience.
A 15-second vertical tour of an apartment in Nairobi, based on the two reference images: enter through the bright living room, glide into the fitted kitchen, then open onto the balcony overlooking the trees. Smooth eye-level camera moves. Sound: calm ambience, soft music. No on-screen text.
An elegant 12-second invitation: a reception hall decorated with white flowers and gold fabric, candles light up one by one, the camera glides toward a long set table, petals fall in slow motion. Sound: soft piano and murmuring guests. No text or subtitles.
Reuse the camera movement from the reference video (slow push-in, then a slight rise) to film the restaurant from the reference image at lunchtime: full tables, waiters moving around, steaming dishes. Sound: busy dining room, clinking cutlery.
Deepnia’s other video models, to pick the right one for your project.
| Model | Length | Resolution | Aspect ratios | Best for |
|---|---|---|---|---|
| Seedance 2.0This model | 4–15 s | 480p–1080p | 16:9 · 9:16 · 1:1 · 3:4 · 4:3 · 21:9 | Realistic videos with sound, in three quality levels. |
| Seedance 2.5 | 4–30 s | 480p–720p | 16:9 · 9:16 · 1:1 · 3:4 · 4:3 · 21:9 | Videos up to 30 s with up to 16 references. |
| Kling 3.0 | 3–15 s | Set by the mode | 9:16 · 1:1 · 16:9 | A shot-by-shot editor, persistent characters and optional sound. |
| Gemini Omni | 4, 6, 8 or 10 s | 720p–4K | 9:16 · 16:9 | 4–10 s clips up to 4K with 7 reference images. |
| HappyHorse | 3–15 s | 720p–1080p | 9:16 · 1:1 · 16:9 · 4:3 · 3:4 | Editing an existing 3–60 s video. |
| Grok Video | 6–30 s | 480p–720p | 9:16 · 1:1 · 16:9 · 2:3 · 3:2 | 6–30 s videos from text or one image. |
They are the three official models in the series: Seedance 2.0 mini, 2.0 fast and 2.0. According to ByteDance, Standard aims for the best quality, Fast balances cost and speed, and Mini offers the best cost-performance. The Standard variant goes up to 1080p.
From 4 to 15 seconds, 8 seconds by default. That is also ByteDance’s official range.
Yes, always: voices, sound effects and music are generated with the picture. Describe the sounds you want at the end of the prompt.
480p or 720p with the Mini and Fast variants, to test and produce quickly; up to 1080p with Standard, for the final cut. Videos come out in 480p by default.
Up to 7 media on Deepnia, including up to 3 videos. ByteDance recommends 4 or 5 well-chosen items rather than the maximum.
Yes: in Reference mode, add the video whose movement you like (up to 3 videos) and say in the prompt what to reuse, for example “reuse the tracking shot from the reference video”.
Yes, with Keyframe mode: the start frame is required, the end frame optional.
Seedance 2.0 for 1080p (Standard variant) or quick tests in Mini. Seedance 2.5 for videos up to 30 seconds and more references.
The video tool opens with Seedance 2.0 already selected. The exact cost shows before every creation; you sign up when you generate.
