Andrew Adams · Co-Founder & Operations at Wireflow · Seedance 2.5
ByteDance's Seedance 2.5 video model generates 30 seconds in a single pass at native 4K, with audio built in the same latent space and up to 50 reference inputs.
Free to build · no credit card
We've run 750+ seedance 2.5 generations internally while building Wireflow and identified the three factors that separate high-quality AI outputs from generic ones — and built them directly into this workflow.
What Seedance 2.5 Is
Seedance 2.5 is ByteDance's video generation model, unveiled in Beijing on June 23, 2026, with a public launch window that opened on July 3, 2026 as an enterprise beta rolling out through Dreamina and Jimeng. ByteDance released nothing between 2.0 and 2.5, so 2.5 is the direct successor to Seedance 2.0 and any spec sheet for an in-between version describes a release that never shipped.
The headline change is length. Seedance 2.5 generates a full 30 seconds in a single pass, so there is no stitching and no seam where two clips join. Output is native 4K at 10-bit color depth, and audio is generated in the same latent space as the video rather than added afterwards, which is what keeps sound and picture in sync natively. It accepts up to 50 multimodal reference inputs, including images, audio clips, 3D white-box models, and style references, roughly four times what Seedance 2.0 allowed. Prompt adherence is about 20 percent better than 2.0, and subjects and characters stay consistent across cuts inside the 30-second window. Pricing has not been announced.
Seedance 2.5 Capabilities
30 Seconds in One Pass
Generates a full 30-second clip in a single pass, so there is no stitching step and no visible seam between segments.
Native 4K, 10-Bit Color
Renders at native 4K with 10-bit color depth, which leaves grading headroom that 8-bit output does not give you.
Audio in the Same Latent Space
Sound is generated in the same latent space as the video rather than bolted on afterwards, so audio and picture line up natively.
Up to 50 Reference Inputs
Takes up to 50 multimodal references: images, audio clips, 3D white-box models, and style refs. That is about four times the Seedance 2.0 cap.
Consistency Across Cuts
Holds subject and character consistent across cuts within the 30-second window, and follows prompts about 20 percent more closely than 2.0.
Chain With Other Models
Wireflow runs video models as nodes next to Flux 2 Pro, Nano Banana 2, and upscalers, so a whole pipeline executes from one endpoint.
More Than Just Seedance 2.5
See What Changed From 2.0
Seedance 2.5 lifts single-pass clip length to 30 seconds, raises output to native 4K, and takes up to 50 reference inputs, roughly four times the 2.0 cap. The Seedance 2.0 page covers the model you can run today.
Call It From Your Code
Seedance 2.5 has no confirmed public API yet. The Seedance 2.5 API page tracks what is known about access and shows the async submit, poll, and retrieve pattern Wireflow already uses for video.
Compare Against Kling 3
Run the same prompt through two video models in parallel branches before committing to one. Kling Video 3 has a free tier and multi-shot clips around 15 seconds, and sits a single node swap away.
One Canvas, Many Models
Seedance is one of several video options on the platform. The AI video generator lets you generate from text or images and swap engines without rewriting any integration code.
Build Visual Video Pipelines
Chain a Seedance video node with image models, upscalers, and LLM prompt steps on a canvas. The node-based video generation editor and the API run the exact same workflow definition.
AI Models Available
Automate Any Workflow
Included in Every Plan
FAQs
Seedance 2.5 is ByteDance's video generation model, unveiled on June 23, 2026 in Beijing with a public launch window that opened July 3, 2026. It generates 30-second clips in a single pass at native 4K, with audio generated alongside the video.
More From Wireflow
Written by
Andrew Adams · Co-Founder & Operations at Wireflow
Runs client operations and content strategy at Wireflow. Works directly with creative teams and agencies to build production AI workflows.
Build Your Video Pipeline on Wireflow
Seedance 2.5 pricing and general API access have not been announced yet. Seedance 2.0 already runs as a node on the Wireflow canvas, wired to image and prompt nodes, with Kling 3 and Veo 3.1 a swap away.