
1 / 3
【Overview】 Reference to Video is an AI video generation workflow that builds new scenes from one or more source assets. Instead of describing everything with text alone, you supply reference images, videos, and audio as creative evidence, and the workflow carries those qualities into the generated shots. This makes it possible to keep a recognizable face, a specific product, a motion style, or an audio-driven beat intact across multiple pieces of content. The problem it addresses is consistency. Generative video often drifts between shots: a character's face changes, a product label shifts, a camera move resets, or timing no longer lines up with the soundtrack. By anchoring generation to reference material, Reference to Video keeps identities, products, motion, camera direction, style, and audio timing aligned from one scene to the next. It suits anyone producing video in sequences rather than one-offs, including episodic creators, AI influencers, product campaigns, anime characters, and fashion sequences. 【Key features】 • Character identity control: Keeps faces, hair, body proportions, and distinctive features recognizable across new scenes. • Product appearance control: Preserves product shape, color, materials, packaging, and label placement in new shots. • Motion and performance control: Uses a short video reference to guide gestures, timing, body movement, and performance energy. • Camera movement control: References a pan, orbit, push-in, handheld move, or other camera behavior that is hard to describe. • Scene and style control: Carries a location, lighting language, palette, or visual world into a newly composed shot. • Audio and timing control: Guides dialogue, voice, music, beats, and action timing on models that accept audio references. • Image, video & audio references: Assigns @Image, @Video, and @Audio sources to separate production jobs within a single generation brief. • Seedance 2.0 Mini: Reference video generation model supporting 480P and 720P output. 【Use cases】 1. Same Character, Different Scenes, Stylized 3D Girl: Game UI Showcase 2. Multi-Character Story 3. Product + Person UGC Ad 4. Anime / OC Character 【Pricing】 Lite: $12/month — 1,500 credits per month, up to 37 videos, 720P, 1080P & 4K output, up to 30s per video, depending on settings Pro: $29/month — 4,000 credits per month, up to 100 videos, 720P, 1080P & 4K output, up to 30s per video, depending on settings Ultra: $89/month — 13,000 credits per month, up to 325 videos, 720P, 1080P & 4K output, up to 30s per video, depending on settings 【Pros and cons】 Pros: - Episodic creators and anyone running serialized characters get the most value, because the workflow is built to hold identity and look steady across many separate generations rather than one clip. - Product marketers and UGC-style advertisers benefit from appearance control that keeps packaging and labels intact, which matters more for commercial shots than for purely artistic ones. - Teams that care about camera language or beat-synced edits will find reference-driven motion, camera, and audio guidance more practical than writing long text prompts. Cons: - Each generation brief accepts up to 15 reference assets (9 images, 3 videos, 3 audio files). - Output resolution and length depend on the selected model: Seedance 2.0 Mini supports 480P/720P; Seedance 2.0 Fast supports 720P; Seedance 2.5 supports 480P/720P up to 30 seconds; MiniMax H3 supports 768P and 2K with direct videos billed by total input plus output duration; Omni Flash 1.1 supports 720P/1080P/4K at 8 or 10 seconds and up to seven weighted reference slots, with one video using two slots. 【FAQ】 Q: Who is Reference to Video designed for? A: It is aimed at episodic creators, AI influencers, product campaigns, anime characters, and fashion sequences — anyone who needs the same character, product, or style to stay consistent across multiple scenes. Q: How much does Reference to Video cost? A: Plans are Lite at $12/month with 1,500 credits, Pro at $29/month with 4,000 credits, and Ultra at $89/month with 13,000 credits. All three list 720P, 1080P & 4K output and up to 30s per video depending on settings. Q: Can I combine image, video, and audio references in one generation? A: Yes. The workflow assigns @Image, @Video, and @Audio sources to separate production jobs within a single generation brief, with up to 15 reference assets in total: 9 images, 3 videos, and 3 audio files.
Lite · $12/month (1,500 credits / month, Up to 37 videos, 720P, 1080P & 4K output) Pro · $29/month (4,000 credits / month, Up to 100 videos, 720P, 1080P & 4K output) Ultra · $89/month (13,000 credits / month, Up to 325 videos, 720P, 1080P & 4K output)
WPS AI is an AI assistant built into WPS Office that helps users write documents, generate presentation outlines and…
Baidu Comate is an AI coding assistant built by Baidu on its ERNIE model. It goes beyond simple autocomplete: it…
UniConverter is an AI media conversion and enhancement tool for video, audio, and images. It combines everyday…
6 similar tools to consider instead of Reference to Video