ByteDanceByteDance·🎬 Video Generation·New

Seedance 2.5 US

private
Try on Venice.ai ↗
Quick reference
Seedance 2.5 US — TLDR
  • 🏢 ByteDance Seed's flagship video family, US-hosted private image-to-video route
  • 🆕 Animates a single still image into motion with native audio
  • 📏 Up to 30 seconds generated in one pass, officially
  • 🔧 Generated clips can be extended twice for longer stories
  • 🎯 ByteDance cites precise reference control and editing capabilities
  • 👁️ Joint audio-video generation with lip sync and camera control
  • ⚡ Smoother, more consistent motion than earlier Seedance releases
  • 🌐 Released September 2026 as a private-deployment variant
💰 Pricing
$1.16 – $46.08
per generation
📅 On Venice since
Sep 28, 2026
4 days ago
Provider

ByteDance is a Chinese internet technology company headquartered in Beijing's Haidian district, with an associated variable-interest entity incorporated in the Cayman Islands. Best known as a consumer internet operator at enormous scale, it has also become a…

Read full profile →
27 models on Venice
21 video · 3 image · 3 inpaint
Since Nov 5, 2025

About this model

Seedance 2.5 US (image-to-video) is the US-hosted, private-deployment route of ByteDance's Seedance 2.5 video family, released in September 2026. It takes a still image as the first frame and generates motion, camera work and synchronized audio from it, sitting alongside its prompt-driven sibling Seedance 2.5 US and the multi-asset Seedance 2.5 US R2V.

ByteDance describes Seedance 2.5 as a next-generation audio-video joint generation model built for 30-second storytelling, with precise reference control and editing capabilities; clips can be produced up to 30 seconds in a single generation and extended twice, with motion the company says is smoother and more consistent and visuals more realistic than before. That single-pass length is the main difference relative to the earlier Seedance 2.0 and Seedance 1.5 Pro lines in this catalog.

Functionally this variant matches the general-availability Seedance 2.5 image-to-video model, differing in deployment: a US-region private endpoint rather than the shared basic route. Typical uses are animating product shots, portraits and concept frames into longer continuous shots where the input subject must hold its appearance for the whole clip.

No technical report for Seedance 2.5 is listed on ByteDance's Seed site, so parameter counts, architecture details and independent benchmark results are not available; the description above reflects the company's own launch material.

This About section is AI-generated from public sources (Claude Opus 5), with no human editing. It may contain inaccuracies — verify critical details against the sources listed above.

Data sources: Venice API · HuggingFace · Wikipedia — enrichment updated 2d ago