Seedance Pro 1.5 by ByteDance — the only Seedance model with native audio generation and cinematic camera control. Pan, zoom, dolly, orbit — all in one generation pass. Uncensored I2V on ZenCreator.
The only Seedance model that generates audio in the same pass as video. Ambient sound, cinematic score, and voice-synced output — baked in, no separate pipeline step.
🎬 Camera control prompts
Pan left, slow zoom in, dolly forward, orbit — write camera direction in your motion prompt and the model follows it. The only Seedance variant with cinematic camera direction.
🔓 Uncensored output
Trusted users get unrestricted generation — ByteDance's content filters are disabled on this private deployment. Same creative freedom as Seedance Pro, with audio and camera on top.
🏆 Top of the Seedance line
Seedance Pro 1.5 sits above Pro and Pro Fast in every dimension except speed. 1080p resolution, smooth natural motion, precise subject tracking — the flagship for storytelling.
📽️ Cinematic storytelling
Best for clips that need a film feel — slow push-in toward a face, orbit around a subject, audio rising as the frame widens. Pro 1.5 is the model when the content has an arc.
📱 Multi-aspect ratio
9:16 for Reels and TikTok, 16:9 for YouTube, 1:1 for feed — all at 1080p with audio and camera control intact across every format.
What is Seedance Pro 1.5?
Seedance Pro 1.5 is the top of the Seedance family — ByteDance's flagship video model with two capabilities no other Seedance variant has: native audio generation and camera-control prompting. Where Seedance Pro generates clean, silent video, Pro 1.5 generates video with ambient sound, music, or voice synced in the same generation pass. Where Pro has no camera direction, Pro 1.5 responds to pan, zoom, dolly, and orbit commands written directly in the motion prompt.
The result is a model built for storytelling rather than simple animation. A slow push-in toward a face with cinematic music underneath; an orbit around a product with environmental ambience; a dolly forward through a scene as audio rises — these are the outputs Pro 1.5 is designed for. The underlying video architecture is the same 1080p Dual-Branch Diffusion Transformer as the rest of the Seedance line, extended with an audio diffusion module and a camera-control conditioning layer.
On ZenCreator, Seedance Pro 1.5 runs inside the Image-to-Video tool. Upload a source image, write a combined motion + camera + audio prompt, select Pro 1.5, and generate. Trusted users get unrestricted output. Note that the audio pipeline adds generation time compared to Seedance Pro — budget extra time for 10-second clips.
See Seedance Pro 1.5 in action
Three source images showing motion prompts that combine camera direction and audio context.
You don't need audio or camera control — Pro 1.5 is slower and heavier than Seedance Pro. For a clean silent clip without camera movement, Pro generates the same 1080p quality faster.
You're testing ideas quickly — for draft screening, use Seedance Pro Fast. Pro 1.5's audio pipeline adds significant generation time per clip.
Your camera prompt is vague — camera control in Pro 1.5 requires precise language ("slow dolly forward 30%", not "move closer"). Vague camera instructions degrade the result; if you're not writing a specific camera move, skip it and use Pro instead.
How to get started
1
Upload your photo
2
Write motion + camera prompt
Stylish night car interior scene, a confident blonde woman lounges across the white leather back seat of a luxury car, wearing a denim tube top, red cargo pants, and white sneakers. Blue ambient lighting glows through the interior, city lights blur outside the window. She lazily shifts her legs across the seat in a relaxed playful motion, settling deeper into the seat with effortless attitude. Right after it pops, she gives a short playful laugh with a confident mischievous expression. Smooth realistic body motion, natural leg movement, subtle laugh, no abrupt movement, no exaggerated acting, cool luxury street aesthetic, modern Instagram reel energy. Camera makes a slow cinematic handheld push-in, keeping her face, legs, and the glowing car interior beautifully framed.
How to write motion + camera prompts for Seedance Pro 1.5
Seedance Pro 1.5 responds to three prompt layers simultaneously: subject motion, camera movement, and audio context. Write each layer separately in your prompt for the cleanest result.
🎬 Camera direction
Name the move explicitly — the model won't invent camera movement unless you ask. Vague phrasing ("move closer") gives inconsistent results.
USE THESE:
slow dolly forward · zoom in to 70% · pan left · orbit clockwise · static camera
🎵 Audio context
Describe the audio you want, not the sound effect name. Mood-based descriptions work better than genre labels for Pro 1.5's audio model.
USE THESE:
ambient score rises softly · distant crowd hum · environmental ambience · no music
🚶 Subject motion
Keep subject motion simple when combining with camera movement. Complex body motion + camera move = higher chance of artifacts. Separate concerns across two layers.
WHEN MOVING CAMERA:
Subject holds still · subtle hair movement · no subject movement
⚡ Prompt order
Lead with camera move, then subject motion, then audio. The model weights earlier tokens more — camera instruction first means the camera actually moves.
TEMPLATE:
[Camera]. [Subject]. [Audio].
Tip: If your prompt doesn't include a camera keyword, Pro 1.5 defaults to static camera — same as Seedance Pro. You only get camera movement when you ask for it explicitly.
Bottom line
Seedance Pro 1.5 is for creators who need the full cinematic toolkit in one generation pass — 1080p output, camera direction, and audio baked in without a separate pipeline. It's the most capable (and the slowest) Seedance model on ZenCreator. Draft on Seedance Pro Fast, finalize on Seedance Pro, and reach for Pro 1.5 when the clip needs a camera move or a soundtrack.
Available in
Image-to-Video
Upload a source image, write a combined motion + camera + audio prompt, pick Seedance Pro 1.5, generate.
Seedance Pro 1.5 adds two capabilities Seedance Pro doesn't have: native audio generation (ambient sound, music, or environmental audio baked into the video at generation time) and camera-control prompting (pan, zoom, dolly, orbit written directly in the motion prompt). The base video quality is the same 1080p output, but Pro 1.5 is slower because of the audio pipeline.
Slow dolly forward/backward, zoom in/out, pan left/right, orbit clockwise/counterclockwise, tilt up/down, and static camera. Write the move name explicitly — the model defaults to static if no camera keyword is present. Combine two moves only if they're physically compatible (slow pan right + slight tilt up works; dolly forward + orbit simultaneously often doesn't).
Seedance Pro 1.5 generates ambient audio based on the scene context and any audio description in your prompt. This includes ambient scores, environmental sounds (crowd hum, nature), and background music. It doesn't generate singing or speech — for synced voice or lipsync, use the Lipsync tool on a clip generated without audio.
For trusted users on ZenCreator, yes. Seedance Pro 1.5 runs on a private ByteDance deployment with content filters disabled — same access level as Seedance Pro and Pro Fast. Contact support if you need trusted user access.
The audio diffusion module runs as an additional pipeline stage alongside the video diffusion process. For 10-second clips the generation time is noticeably longer than Seedance Pro. If you need quick output without audio or camera control, use Seedance Pro instead.
Yes — any video editor (DaVinci Resolve, Premiere, even QuickTime) can strip the audio track. If you want to replace the audio with a custom track or voiceover, generate without an audio prompt first to get a silent clip, then add audio separately in your editor or via the Lipsync tool.
Sources
ByteDance Seed lab — Seedance 1.5 model technical report: seed.bytedance.com
