Seedance25Try on ponpon
Released June 11, 2025 — the original Seedance

Seedance 1.0

Seedance 1.0 was ByteDance's first publicly released Seedance AI video model, announced at the 2025 Volcano Engine Spring FORCE conference and shipped through the Doubao app and Dreamina/Jimeng. It generates text-to-video and image-to-video clips of roughly 2–12 seconds at up to 1080p/24fps — video only, with no native audio.

It has since been superseded — Seedance 1.5 pro added native audio, and Seedance 2.0 is the version most people use today (full history in the version timeline) — but 1.0 still matters as the model that started the line, and it's still served through APIs.

Text-to-video & image-to-video

Describe a scene from scratch, or animate a still image — the same two modes the whole Seedance line still uses today.

Multi-shot storytelling

Native multi-shot narrative within one generation, plus first/last-frame control to pin how a clip starts and ends.

Up to 1080p at 24fps

Clips of roughly 2–12 seconds at 480p, 720p or 1080p — solid for its time, though silent: audio only arrived with 1.5 pro.

Lite & Pro variants

Seedance 1.0 shipped as Lite (faster, cheaper) and Pro (higher quality) — the Pro variant is the one still served as doubao-seedance-1.0-pro.

Is Seedance 1.0 still worth using?

For most people, honestly, no — Seedance 2.0 generates native audio, accepts multimodal references, runs longer clips, and simply looks better. The realistic case for 1.0 today is API cost: developers who need short, silent clips at the lowest rate can still call doubao-seedance-1.0-pro on Volcano Engine / BytePlus ModelArk, or the Seedance v1 endpoints on fal — typically cheaper than 2.0 (rates change, so check the provider). For everything else, start with the current version. See the Seedance API guide for the full provider list.

Frequently asked questions

What is Seedance 1.0?

Seedance 1.0 is ByteDance's first publicly released Seedance AI video model, launched on June 11, 2025 at the Volcano Engine Spring FORCE conference. It generates text-to-video and image-to-video clips of roughly 2–12 seconds at up to 1080p/24fps, with multi-shot storytelling and first/last-frame control.

Is Seedance 1.0 still available?

Yes — mainly through APIs. ByteDance's Volcano Engine / BytePlus ModelArk still serves doubao-seedance-1.0-pro, and partners like fal expose Seedance v1 endpoints. Consumer apps such as Dreamina have moved to newer versions, so for hands-on use most people are on Seedance 2.0 today.

What's the difference between Seedance 1.0 and 2.0?

Seedance 1.0 generates silent clips (no audio) of 2–12 seconds at up to 1080p. Seedance 2.0 (February 2026) is a unified multimodal model: it accepts text, image, audio and video references, generates native audio including lip sync, produces ~15-second clips, and a later update added native 4K. For most work, 2.0 is the better choice.

Is there a Seedance 1.5?

Only "Seedance 1.5 pro", released December 16, 2025 — there is no separate non-pro Seedance 1.5. Its headline feature was native synchronized audio and video, which set the direction for Seedance 2.0.