At its Volcano Engine FORCE conference, ByteDance showed a model that renders a continuous 30-second clip without stitching and accepts up to 50 reference inputs — in enterprise beta now, public in early July.
2026-06-24 · by Moe Ameen
ByteDance unveiled Seedance 2.5, the next version of its text- and image-to-video model, at its Volcano Engine FORCE conference on June 23, 2026. The headline capability is clip length: Seedance 2.5 generates a continuous 30-second video in a single pass — no post-stitching — and can carry scene changes and tempo shifts inside that one render. For most of the year, leading models had been limited to roughly a 4-to-15-second band, so breaking the 30-second mark in a single generation is a notable jump.
The model also widens its inputs. Seedance 2.5 accepts up to 50 reference files in one request — images, audio, video, style references, and 3D models — up from 12 in Seedance 2.0, which gives much tighter control over look, motion, and consistency across a shot. ByteDance additionally describes native 4K output and post-generation editing that preserves the clip's visual style. Seedance is the engine behind ByteDance's creative apps, including Dreamina, the Doubao assistant, and CapCut, and is distributed to businesses through Volcano Engine, ByteDance's cloud platform.
Two things are worth keeping straight. Seedance 2.5 is not generally available yet: ByteDance framed it as a global enterprise beta, with a public launch targeted for early July 2026. And there is no published independent benchmark for 2.5 — any score circulating now is unverified. The credible quality signal is the predecessor: Seedance 2.0 has led the independent Artificial Analysis Video Arena for both text-to-video and image-to-video. Treat specific resolution, length, and pricing figures as a snapshot until the public release.
There are two ways to act on this, and the first one you can do today. A model launch from a player the size of ByteDance is a story your audience cares about — so publish your take before the recap wave crests. Drop your point of view on the 30-second milestone into Kompozy as a source, and it spins that one angle into a blog explainer, a carousel breaking down what single-pass 30-second generation changes, short captioned clips, and platform-native posts in your voice through the Persona Brief — then schedules and publishes the set across TikTok, Reels, YouTube Shorts, Facebook, LinkedIn, X, Pinterest, and Threads from one queue. Being early on a launch like this is how a single take becomes a week of content.
The second way lands once Seedance opens up in early July. When you can generate a 30-second clip, that clip is finally long enough to be a short by itself — and Kompozy is the layer that finishes it: branded captions, per-platform reframing to 9:16, 1:1, or 16:9, a hook overlay through HyperFrames on the silent first second, then scheduling and publishing across all nine platforms. Kompozy also generates the video Seedance can't — a HeyGen talking-head Persona Short, an avatar composited into a brand-exact template, a listicle video over a portrait clip — so your recurring on-brand identity stays put while you swap in Seedance renders as accent footage. The model makes the scene; Kompozy makes the posts.
It is ByteDance's text- and image-to-video model, unveiled at the Volcano Engine FORCE conference on June 23, 2026. Its standout feature is generating a continuous 30-second clip in a single pass without stitching, while accepting up to 50 reference inputs. It powers ByteDance apps like Dreamina, Doubao, and CapCut.
At announcement it was a global enterprise beta, with a public launch targeted for early July 2026. Access, limits, and pricing were still settling, so confirm current availability through Volcano Engine.
Not yet. There is no published independent benchmark for 2.5, so any score circulating now is unverified. The credible signal is that the predecessor, Seedance 2.0, has topped the independent Artificial Analysis Video Arena for text-to-video and image-to-video.
Seedance generates the clip but does not publish it. Bring the export into Kompozy to add branded captions, reframe per platform, stack a hook overlay, fan it into a carousel and captions in your voice, and schedule and publish across TikTok, Reels, YouTube Shorts, X, LinkedIn, and more from one queue.