Vidu Q3 / Reference to Video
Base reference-to-video model built for narrative video creation. Supports native audio-video generation and multi-character dialogue. Delivers robust character consistency and scene coherence across up to 16-second clips
Vidu Q3Subject-Reference Video with Audio
Vidu Q3 generates reference-conditioned video at 540p–1080p with optional native audio, audio type selection, and watermark position control.
At a Glance
Vidu Q3 Architecture
Subject-aware multimodal video for production.
Subject Reference Conditioning
Feed subjects so identity stays locked across generated motion variants.
Optional Native Audio
Enable audio generation with type selection — all, speech only, or sound effects only.
540p to 1080p
Draft at 540p, ship social at 720p, and promote heroes to 1080p.
Watermark Position Control
Optional watermark with selectable corner positions (1–4) and custom watermark URL.
How It Works
From subject reference to finished clip.
Prepare Subjects
Collect reference subjects that define identity.
Write the Motion Brief
Describe the action and camera language.
Set Resolution & Audio
540p/720p/1080p; optional audio type.
Watermark & Deliver
Optionally place a watermark and pull the finished clip.
Vidu Q3 Domains
Where subject identity and sound ship together.
Character Motion
Reference-conditioned character clips.
Social with Sound
Native audio for Reels and TikTok.
Product Film
Reference-conditioned product motion.
Branded Watermarks
Corner position and custom watermark URL.
Prompt Tips
Cleaner Vidu Q3 output.
Subject references work best when each image has a single clear job.
If audio is on, name environmental cues so sound matches picture.
Explore motion cheaply, then promote to 1080p for finals.
Vidu Q3 Quickstart
Subject-reference video generation.
curl -X POST "https://api.powertokens.ai/v1/videos" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "viduq3",
"prompt": "Cinematic tracking shot through a neon cyberpunk city in heavy rain.",
"seconds": "5",
"size": "1080p",
"ratio": "16:9",
"resolution": "1080p",
"duration": 5,
"aspect_ratio": "16:9",
"audio": true
}'Technical Specifications
Confirmed parameters and runtime execution protocols.
Pricing Details
The actual billing for this model is dynamically calculated based on the specific parameters passed in your API request. Below are the specific combinations and their corresponding pricing:
| Modality | Credits | Price (USD) |
|---|---|---|
| Reference to Video540PPeak Shifting | 19/ Second | $0.019 |
| Reference to Video540P | 34/ Second | $0.034 |
| Reference to Video720PPeak Shifting | 29/ Second | $0.029 |
| Reference to Video720P | 57/ Second | $0.057 |
| Reference to Video1080PPeak Shifting | 34/ Second | $0.034 |
| Reference to Video1080P | 72/ Second | $0.072 |
Models from the Same Channel
Explore complementary models and alternative versions from the same provider channel.
Vidu Q3 Pro
viduq3-pro
Flagship video model supporting text-to-video, image-to-video, and start/end-frame workflows. Generates up to 16-second clips with native audio sync and storyboard capabilities. Available in 540p–1080p resolutions with premium motion dynamics
Vidu Q3 Turbo
viduq3-turbo
Optimized for speed, generating 1–16 second video clips from text or images with faster inference than the Pro variant. Supports 540p–1080p output, balancing generation quality with reduced latency for rapid iteration
Vidu Q3 Pro Fast
viduq3-pro-fast
High-efficiency variant optimized for cost and generation speed, ideal for high-volume creative pipelines. Supports 1–16 second image-to-video at 720p or 1080p, offering the most affordable per-second pricing in the Q3 series
Recommended Related Models
Explore complementary video and multimodal models with your unified API key.
Seedance 2.5
dreamina-seedance-2-5-260628
ByteDance's latest flagship video generation model, built for longer-form storytelling and production-ready output. Generates up to 30 seconds of continuous, cinematic video with native audio sync in a single pass . Accepts up to 50 multimodal references (images, videos, audio, character sheets, storyboards) for precise scene, character, and motion consistency . Features localized region editing to fix specific areas without full regeneration……
Wan 3.0
wan3.0-video
Wan3.0-Video is an all-in-one video generation model unified support for multiple creative capabilities, including reference, editing, replication, and driving. It generates videos up to 30 seconds with omni-modal reference, and can parse files, web pages and complex images. With production-grade character consistency and lifelike visuals and sound, it delivers an immersive audiovisual experience.
Seedance 2.0 Mini
dreamina-seedance-2-0-mini-260615
Lightweight, cost-efficient video model from ByteDance, optimized for speed and high-volume content creation. Supports text-to-video, image-to-video, and reference-based generation with up to 12 references (6 images, 3 audio, 3 video). Delivers faster generation and lower credit consumption than Seedance 2.0, with strong motion quality and character consistency. Ideal for social media content, product videos, AI short dramas, and rapid creative iteration
Seedance 2.0
dreamina-seedance-2-0-260128
Generate videos from reference images, videos, and audio; edit videos; extend videos; generate videos from start and end frames
Frequently Asked Questions
Everything you need to know before integrating this model.
Start Building with Vidu Q3 Today
Create an account in seconds to receive 100 free credits and start generating immediately. No credit card or upfront contract required.