Seedance version history: from 1.0 to 2.5 Flash

Quick answer: Seedance is ByteDance’s video model family. Seedance 1.0 pro launched on June 11, 2025; Seedance 1.5 pro added generated sound in December 2025; Seedance 2.0 brought multimodal references in February 2026; Seedance 2.5 raised a take to 30 seconds on July 31, 2026; and Seedance 2.5 Flash, the faster and cheaper model, followed on November 12, 2026. SceneMixer offers the Seedance 2.0, 2.5 and 2.5 Flash families as video tiers.
Still from “The Last Envelope”, an AI short drama generated end-to-end by SceneMixer (6 characters, 1 episode)
SampleThe Last EnvelopeDigital HD1 episodeProduced end-to-end by SceneMixer from the scriptWatch the episodes

Seedance 1.0 pro — June 11, 2025

Announced at Volcano Engine’s FORCE conference alongside the Doubao 1.6 language models. It took text and images and produced 1080p video with cuts between several shots in one clip, and its launch price — about 3.67 yuan for a five-second 1080p clip — was pitched as the lowest in the industry. No sound.

Seedance 1.5 pro — December 16, 2025

The first Seedance to generate sound with the picture: speech, ambience, effects and music in one pass, with lip sync and support for several languages and Chinese dialects. ByteDance’s own post is dated December 16; the API followed on Volcano Engine Ark about a week later.

Seedance 2.0 — February 12, 2026

A unified architecture for joint audio-video generation that accepts four kinds of input: text, images, audio and video. This is the release that made reference-driven video mainstream — a face from one image, a camera move from a clip, a voice from a recording — and it drew complaints from film studios within days of launch.

Seedance 2.5 — July 31, 2026

Thirty-second native clips and up to 50 multimodal references per request, with local editing of part of a clip. For series work the longer take is the change that matters: a dialogue scene needs half as many joins.

Seedance 2.5 Flash — November 12, 2026

The fast model of the 2.5 generation: about twice the speed and roughly 34% less per second by ByteDance’s figures, 30-second takes with sound, 480p and 720p, up to 20 references. The API opened on November 15. Seedance 2.5 Flash in detail

VersionReleasedWhat it added
Seedance 1.0 proJune 11, 2025Multi-shot 1080p video from text or an image
Seedance 1.5 proDecember 16, 2025Sound generated with the picture; lip sync
Seedance 2.0February 12, 2026Text, image, audio and video as inputs
Seedance 2.5July 31, 202630-second takes; up to 50 references
Seedance 2.5 FlashNovember 12, 2026About twice the speed, about 34% cheaper; 480p and 720p

Which versions SceneMixer offers

  • Seedance 2.5 Flash: Seedance 2.5 Flash (720P) and Seedance 2.5 Flash · 480P (480P), takes up to 30 seconds.
  • Seedance 2.5: Seedance 2.5 Pro (1080P), Seedance 2.5 Lite (720P) and a 480P tier, takes up to 30 seconds. Seedance 2.5 on this site
  • Seedance 2.0: Seedance 2.0, Seedance 2.0 Fast and Seedance 2.0 · 480P, takes up to 15 seconds. Seedance 2.0 on this site

The tier is one setting in the editor, so a project can move between generations segment by segment.

Sources

Dates are release or announcement dates as given by those sources, checked on November 16, 2026.

FAQ

What is the latest Seedance version?

Seedance 2.5 Flash, released on November 12, 2026, is the newest model; Seedance 2.5 is the flagship.

When was Seedance 2.0 released?

February 12, 2026, by ByteDance's own announcement.

When was Seedance 2.5 released?

July 31, 2026.

Which Seedance versions can I use in SceneMixer?

SceneMixer offers Seedance 2.0, Seedance 2.5 and Seedance 2.5 Flash as video tiers.

Which Seedance version first had sound?

Seedance 1.5 pro, in December 2025.

Is Seedance 2.5 Flash a newer generation than 2.5?

No. It is a faster, cheaper model within the 2.5 generation.

Who makes Seedance?

ByteDance; it is sold through Volcano Engine in China and BytePlus elsewhere.

Use the newest Seedance on a whole story

Paste a script and pick a Seedance tier for each segment.

Start a project

Updated November 16, 2026