Seedance 1.5 Pro
ByteDance's joint audio-video foundation model — native sound, multilingual lip-sync, and cinematic camera work in a single 12-second take.
# image_url: the first frame, as a URL or a data: URL
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "seedance-1-5-pro-image-to-video-basic",
"prompt": "The camera slowly pushes in as mist drifts across the scene",
"image_url": "https://example.com/start-frame.jpg",
"duration": "5s",
"resolution": "720p"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Overview
What is Seedance 1.5 Pro
Seedance 1.5 Pro is ByteDance's audio-visual video generation model, released in December 2025. Built on a dual-branch Diffusion Transformer, it generates video with native synchronized audio — dialogue, voices, and sound effects — plus multilingual lip-sync, cinematic camera control, and clips up to 12 seconds at 1080p.
Using it anonymously on Venice
On Venice, Seedance 1.5 Pro runs under the anonymized privacy tier: your prompts are not stored, profiled, or fed into anyone's training pipeline. You get the uncensored, permissionless version — pay per clip from $0.14 instead of a subscription, with both text-to-video and image-to-video modes available. No account history ties your creative work back to you.
Specifications
Datasheet
- Maker
- ByteDance (Seed team)
- Modality
- Text-to-video and image-to-video, both with native joint audio generation
- Open weights
- No — proprietary
- License
- Proprietary
- Clip lengths
- 4s – 12s
- Resolutions
- 1080p, 720p, 480p
- Mode
- image-to-video
- Audio
- Yes
- Prompt limit
- 3,500 chars
- Input images
- One still image as the animation source (image-to-video mode)
- Released
- December 16, 2025
- Architecture
- Dual-branch Diffusion Transformer with a cross-modal joint module; RLHF post-training and a 10×+ inference acceleration framework
- Privacy on Venice
- Anonymized — prompts not stored
- Available on Venice since
- Mar 2026
Assessment
Strengths and limitations
- Native joint audio-visual generation: dialogue, voices, and spatial sound effects are produced in the same pass as the video, not bolted on afterward.
- Precise multilingual and dialect lip-syncing with motion alignment, one of the strongest dialogue engines in its class.
- Film-grade cinematography: complex camera movement from close-ups with subtle facial emotion to full cinematic shots.
- Strong narrative coherence: it auto-fills story intent and keeps characters' emotions, expressions, and actions consistent across a clip.
- Fast: ByteDance's acceleration framework reportedly boosts inference speed by over 10× (per the official paper abstract).
- First-frame and last-frame interpolation gives you start/end frame control over a shot.
- Closed and proprietary: no open weights, so no self-hosting or fine-tuning; if you need open weights, Wan 2.7 on Venice is the closest alternative.
- Clips max out at 12 seconds, shorter than the 15–16 seconds some Venice-hosted rivals offer.
- No reference-asset input: it won't take a library of images, video, or documents as character/style references the way some competitors do.
- Third-party blind testing has found rivals such as Kling v2.6 ahead on motion realism and character consistency, even where Seedance wins on cost and text rendering.
Use cases
What it is good for
- 01Short dramas and dialogue scenes where characters need to speak with accurate lip-sync.
- 02Advertising creative — a single generated take with visuals, voice, and sound design.
- 03Social media clips with strong storytelling and emotional expression.
- 04Animating a product photo, portrait, or illustration into motion via image-to-video.
- 05Cinematic shots with directed camera movement, from intimate close-ups to sweeping full-shots.
Prompting
Getting better results
Put spoken dialogue in quotes and name the language or accent — the model generates the voice and lip-syncs it natively.
Direct the camera explicitly: "slow dolly-in to a close-up" or "full-shot, wide angle" — Seedance 1.5 Pro follows cinematography instructions well.
Pick the shortest clip length that works (4s–12s) — price scales with duration, so iterating at 4s in 480p is the cheapest loop.
Draft at 480p, then re-render the winning prompt at 1080p to keep costs down while iterating.
For image-to-video, start from a still with clear composition and describe only the motion you want added, not the whole scene.
Describe sound design in the prompt ("rain on a tin roof, distant thunder") — spatial effects are generated alongside the visuals.
Samples
Sample outputs
Generated on Venice with our standard prompt suite — the same prompts we run through every model of this type, so you can judge it like-for-like.
Slow aerial drone shot gliding over a misty mountain valley at golden hour, sunlight piercing clouds onto a winding river, ancient pine forests on either side, ultra-smooth motion, professional color grading, atmospheric haze, 4K
Calm ocean waves rolling onto a black-sand beach at sunrise, golden light on wet sand, foam dissolving into the shore, a single silhouetted figure at the waterline, smooth continuous forward push, serene cinematic atmosphere
A slow tracking shot through a rain-soaked Tokyo street at night, neon reflecting in puddles, steam rising from a food cart, a person with a translucent umbrella, shallow depth of field, teal-and-orange grade, smooth steady camera
Alternatives
How it compares
| Model | Best for | Max duration | Max resolution | Native audio |
|---|---|---|---|---|
| Seedance 1.5 Pro | Dialogue & audio-visual storytelling | 12s | 1080p | Yes |
| Grok Imagine 1.5 | Fast stylized clips | 15s | 1080p | Yes |
| Wan 2.7 Enhanced | Open-weights video + audio | 15s | 1080p | Yes |
| Vidu Q3 | Longest single takes | 16s | 1080p | Yes |
Seedance 1.5 Pro is the right pick when the clip has to talk — dialogue, lip-sync, and synchronized sound in one generated take. If open weights or longer single takes matter more, Wan 2.7 Enhanced or Vidu Q3 are the alternatives.
API
Call it from your code
Venice exposes this model through the REST API. Queue a generation with the model id and the inputs it needs.
# image_url: the first frame, as a URL or a data: URL
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "seedance-1-5-pro-image-to-video-basic",
"prompt": "The camera slowly pushes in as mist drifts across the scene",
"image_url": "https://example.com/start-frame.jpg",
"duration": "5s",
"resolution": "720p"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Pricing
What it costs on Venice
Pay per clip on Venice — price scales with resolution and duration (4s–12s), from $0.14.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
FAQ
Frequently asked questions
Seedance 1.5 Pro is ByteDance's audio-visual video generation model, officially released on December 16, 2025. It uses a dual-branch Diffusion Transformer to generate video and native audio together — dialogue, voices, and sound effects with multilingual lip-sync — in clips up to 12 seconds at 1080p.
Venice charges per clip, with price scaling by resolution and duration. A 4-second clip starts at $0.14 in 480p, $0.31 in 720p, and $0.69 in 1080p, with longer clips up to 12 seconds priced proportionally. No subscription is required.
It is neither free-software nor open source — ByteDance keeps the weights proprietary, so it cannot be self-hosted or fine-tuned. On Venice you pay per clip rather than a flat fee, and new accounts can try it with welcome credits before paying.
Yes. The seedance-1-5-pro-image-to-video-basic variant animates a still image into motion, accepting one image as the starting frame and generating clips from 4 to 12 seconds at 480p, 720p, or 1080p with native audio. A separate text-to-video variant generates purely from a written prompt.
Both Venice variants generate clips from 4 to 12 seconds at 1080p, 720p, or 480p, with native audio on every clip. Prompts can run up to 3,500 characters.
Seedance 1.5 Pro is generally considered stronger on dialogue, lip-sync, and audio-visual synchronization, and third-party blind testing in early 2026 found it roughly 57% cheaper than Kling v2.6 while winning on text rendering. Kling won that same test on motion realism and character consistency, so the choice depends on whether your scenes talk or move.
Yes — that is its headline feature. It is a native joint audio-visual model: voices, dialogue, and spatial sound effects are generated in the same pass as the video, with multilingual and dialect lip-syncing and motion alignment, rather than audio being added in a separate step.
Yes. Venice runs it under the anonymized privacy tier — prompts are not stored, profiled, or used for training, and generations are not tied to a personal history. You can use it in the Venice app or through the Venice API without building a traceable creative record.
No. Unlike some rivals, it does not accept a library of reference assets — its image-to-video mode takes a single still image as the starting frame, and text-to-video works purely from the prompt. There is no multi-asset reference or video-reference input.
Use Seedance 1.5 Pro anonymously
Venice does not store your prompts. Chat history stays in your browser.