Reeldo AIReeldo AI

Reference to Video AI Generator

Paste a link or upload refs — see popular use cases below, then try it in the tool.

Try the generator

Model

Upload references

@image1 · @video1 · @audio1

Images@image1

Need a reference image? Create in Frame Lab

Reference Images

Up to 9 — character, scene, or style

Video@video1

Paste TikTok / Reels / Facebook link, or upload · 2–10s

Audio@audio1

Upload MP3/WAV · ≤15s total

Remix presets

0/5000
16:9 · 5s · 720p

Popular reference-to-video use cases

Real situations creators run into — what you upload, what you get back, and which @tags to use.

Use case 01 / 06

Vertical TikTok dance clip — motion and camera reference

@video1

Studio portrait — character identity reference

@image1

Output

Portrait character performing TikTok dance motion in AI video output

Ride a TikTok trend with your face

Scenario
You found a dance blowing up on TikTok and want your character — not the original creator — in that exact motion.
Result
Your portrait performs the same choreography and camera moves, ready to post without filming.
Settings
@video1 motion@image1 character

Prompt template

Use @video1 as the motion and camera reference — same choreography, pacing, and camera path. @image1 defines the character's face, hair, and identity throughout the clip. Replace the original dancer with @image1 performing the exact moves from @video1. Vertical 9:16, social-native framing, native sound matching the motion. No on-screen text, logos, or watermarks.

Use case 02 / 06

Vertical UGC ad clip — camera and hand-gesture reference

@video1

Product packshot on white — SKU identity reference

@image1

Output

Product hero following UGC camera motion in vertical AI ad

Drop your product into a proven UGC ad

Scenario
You have a packshot and a competitor’s vertical ad that converts — you need your SKU in that same hand-held, social-native format.
Result
Your product follows the reference camera path and gestures in a 9:16 clip you can run as paid social.
Settings
@video1 camera@image1 product9:16

Prompt template

Use @video1 for camera motion, hand gestures, and social-native pacing. Feature the product in @image1 as the hero — accurate shape, color, and branding. Scene: [optional lifestyle setting]. Vertical 9:16, clean product read, scroll-stopping energy. No on-screen text or logos.

Use case 03 / 06

Portrait photo — face and outfit identity reference

@image1

Reference clip still — wave and walk performance

@video1

Output

Same character performing reference clip action in AI video

Make your character copy a reference performance

Scenario
You have one hero photo of your mascot or influencer and a clip whose wave, walk, or gesture is perfect — but the actor is wrong.
Result
The same identity from your photo performs the reference action in a new video.
Settings
@image1 identity@video1 performance

Prompt template

Keep the character identity from @image1 — face, hair, and outfit. Perform the action, gestures, and camera style from @video1. Scene: [where they are]. Consistent likeness across the clip, cinematic lighting, no identity drift. No on-screen text or logos.

Use case 04 / 06

Character portrait — identity reference for new scenes

@image1

Output

Same character in a new café scene with consistent face

Publish episodic content with one locked face

Scenario
You’re building an AI influencer or brand avatar and need episode 2, 3, 4 in new locations without the face drifting.
Result
Each new scene keeps the portrait identity while the environment and story change in your prompt.
Settings
@image1 portraitscene in prompt

Prompt template

@image1 defines the character — same face, hair, and style in every episode. Scene: [describe the new location and story beat for this episode]. Cinematic lighting, stable identity across shots, no face drift between frames. No on-screen text or logos.

Use case 05 / 06

Dancer portrait — performer identity reference

@image1

Audio waveform visualization — rhythm and beat reference

@audio1

Output

Dancer synced to reference audio beat in AI video output

Choreograph or lip-sync to your track

Scenario
You already have the song or voice-over and a performer look — you need motion that hits the beat, not a silent placeholder.
Result
The character moves and lip-syncs in time with your uploaded audio.
Settings
@image1 performer@audio1 rhythm

Prompt template

Sync visual beats and lip-sync to @audio1. @image1 defines the performer — face, outfit, and stage presence. Energetic dance or performance cover, vertical 9:16, motion hits the rhythm of @audio1. No on-screen text or logos.

Use case 06 / 06

Cinematic still — teal-orange color grade and mood reference

@image1

Output

New scene matching reference film look and color grade

Carry a film grade into a brand-new shot

Scenario
You love the color, contrast, and mood of one cinematic still but need a different subject or location for the campaign.
Result
The new scene inherits the reference look — teal-orange grade, grain, lighting — while content follows your prompt.
Settings
@image1 style refscene in prompt

Prompt template

Use the color grade, contrast, grain, and lighting mood from @image1 as the visual style reference. Scene: [new subject or location for this shot]. Match the cinematic look of @image1 while the content follows your scene description. Slow, deliberate camera move. No on-screen text or logos.

How @tags map to your uploads

Upload order in Reference Studio is tag order in your prompt — match the diagram on the left.

@image1

Upload #1

@video1

Upload #2

@audio1

Upload #3

Prompt

@image1 in the scene from @video1

  1. 1

    @image1

    First image you upload — identity, product, or style still.

  2. 2

    @video1

    First video or pasted TikTok / Reels link — motion and camera.

  3. 3

    @audio1

    First audio clip — rhythm, lip-sync, or beat.

Assign roles in one line: who (@image1), motion (@video1), beat (@audio1).

Which model?

Seedance 2.5 for long clips and link import · MiniMax H3 when audio drives the performance.

Seedance 2.5 reference to video — long clips and link import

Seedance 2.5

Up to 30s, TikTok import, multi-ref, up to 4K on Pro.

  • ·Up to 30s · TikTok link import
  • ·Multi-ref · up to 4K (Pro)
Seedance specs
MiniMax H3 reference to video — audio-driven clips

MiniMax H3

Native stereo audio — best for dance, lip-sync, and performance.

  • ·Native stereo audio · up to 2K
  • ·Best for voice, music, performance
H3 specs

Reference video billing = input + output duration · Details on Pricing

Try in the tool above

Learn more

Reference to video = new scene from multiple refs. Image to video = one still animated. Text to video = prompt only.

Reference to video vs image to video vs text to video comparison diagram

Reference to video = new scene guided by refs · Image to video = animate one frame · Text to video = prompt only

Reference to video FAQ

What is reference to video AI?

Reference to video vs image to video — what's the difference?

Which video links can I paste?

Which models support reference to video?

How do @image1 and @video1 tags work?

How many references can I upload?

How are credits calculated with reference video?

Can I use reference to video commercially?

Explore more

Text to VideoImage to VideoSeedance 2MiniMax H3URL to VideoPricingVideo Studio