ByteDance
ByteDance Seedance
Vertical-first motion tuned for feeds and trend-native framing.
Seedance is optimized for how people actually watch — phone aspect, fast reads, motion that still scans when the phone is on mute. Pair with 9:16 in Generate and clipping agents for downstream hooks.
What you get
- Video with lip-synced or driven motion
What you can send
EchoClip matches Generate workflows and the model picker to these slots — not a single upstream default model.
Required (at least one route)
- 1 start image
- Audio / voice track
- Start frame image (image-to-video)
- Text prompt
- Text prompt (text-to-video)
Optional
- Aspect ratio (9:16 typical)
- Duration
- Strength / steps controls
Studio routes in this family
Each fal endpoint has its own input contract. Pick the route that matches the media you have (start frame, FLF, references, audio, etc.).
In: Text prompt · 1 start image · Audio / voice track · Out: Video with lip-synced or driven motionText prompt: required
Example workflows
- Smoke test — Prompt only · Optional: Strength / steps controls
When to use it
- TikTok/Reels/Shorts where vertical composition matters
- Trend-style product demos and quick-turn campaigns
- Volume testing before you invest in a flagship render
Reach for something else when
- Widescreen narrative or presentation decks (prefer 16:9-first models)
- You need frame-accurate FLF storyboards — use Kling FLF flows
EchoClip MCP (Cursor & Claude)
Connect EchoClip in chat to quote credits, pick this model family, and kick off renders without leaving your editor.
Use MCP to estimate_studio_generation for Seedance at 9:16 before a multi-clip campaign.
MCP setup guides — authenticate once, then generate from chat.
Workflows in EchoClip
- 9:16 hooks
- Feed-native variants
- Agents → approve → post
Best for
- Vertical social
- Product teasers
- Fast A/B hooks
Use ByteDance Seedance in Generate
Open the Video studio, pick this model in the picker, and check live credit estimates before you render.