Seedance
Visit Seedance →About Seedance
Seedance is an AI video generation model developed by ByteDance. The current version, Seedance 2.5, was announced July 31, 2026, rolling out first on Jimeng AI and Doubao Pro, with API access through BytePlus ModelArk. Dreamina, ByteDance's international app, began offering it to subscribers in Europe, Asia, the Middle East and South America starting late October 2026; it is not yet available in the US, where BytePlus also does not operate.
Seedance 2.5 accepts up to 50 multimodal references in a single generation request, combining up to 30 images, 10 video clips and 10 audio clips. A single generation pass runs up to 30 seconds, double the 15-second limit of the prior version, and clips can be extended; Dreamina's beta long-video mode reportedly reaches up to 180 seconds. The model also supports joint audio-video generation along with timestamp-level, green-screen, camera-perspective and reference-based editing.
Beyond ByteDance's own apps (Jimeng AI, Doubao, Dreamina, Dola) and the BytePlus ModelArk API, Seedance is also offered through partners including Higgsfield, Runway and DreamBench.
ByteDance has not published a list price for the Seedance model itself. Access is bundled into Dreamina's credit-based subscription plans, shown in local currency with tiers from a lower-credit Basic plan to a higher-credit Ultra plan; using Seedance 2.5 on Dreamina requires a paid subscriber account. Pricing for the BytePlus ModelArk API was not published at time of review.
Outputs carry visible watermarks and C2PA Content Credentials on BytePlus, or invisible watermarks, C2PA and visible AI labels on Dreamina; no free, watermark-free commercial tier is confirmed. The model is also not currently accessible in the US through either Dreamina or BytePlus, and independent documentation remains limited.
Seedance 2.5 suits video creators, marketers and developers outside the US who want long single-pass generation, multimodal input, and built-in extension and editing tools, particularly those already working within Jimeng, Doubao, Dreamina or BytePlus ModelArk.
- Accepts up to 50 multimodal references per generation request, combining up to 30 images, 10 video clips and 10 audio clips
- Single-pass generations run up to 30 seconds, double the prior version's 15-second limit, with extension options reaching up to 180 seconds in Dreamina's beta long-video mode
- Supports joint audio-video generation plus timestamp-level, green-screen, camera-perspective and reference-based editing
- Available across ByteDance's own apps (Jimeng AI, Doubao, Dreamina, Dola) and the BytePlus ModelArk API, plus partners like Higgsfield, Runway and DreamBench
- No published list price for the model itself; Dreamina pricing is shown in local currency and not confirmed in USD
- Using Seedance 2.5 on Dreamina requires a paid subscriber account; no free access to the model is confirmed
- Not available in the US: Dreamina's 2.5 rollout excludes the US and BytePlus does not operate there
- Outputs carry visible or invisible watermarks and C2PA labeling
- Less established brand recognition and third-party documentation compared to Kling AI, Runway, or Sora
Choose Seedance if…
- ✅ You want ByteDance's latest video model with long single-pass generation and built-in editing tools
- ✅ You're outside the US and can access Dreamina, Jimeng AI, Doubao, or BytePlus ModelArk
- ✅ You need multimodal input — text, image, audio and video — combined in one generation request
- ✅ You want extension and reference-based editing beyond simple text-to-video
Leave a Review
Reviews are published after moderation. We don't share your email.