A photo for each character
Separate uploads keep the left and right references easy to follow. Check the role labels or swap the pair before starting.
Nano Banana 2 / Character duet 01
Nano Banana 2 / Character casting
Two character photos. One 10-second Hotel Lobby video.
JPG, PNG, WebP or BMP · Maximum 20 MB
JPG, PNG, WebP or BMP · Maximum 20 MB
First photo: left performer. Second photo: right performer.
Your characters are placed into this prepared performance.
Bring two familiar faces into the Hotel Lobby performance. Choose who appears on each side, select the video format, and create a 10-second duet from the fixed orange-studio clip.
Character pairings in motion
Watch and listen to these example takes for ideas about character contrast, coordinated outfits, and the shared orange stage.
A pair of grandmothers gives the same stage a different presence.
An everyday character and a theatrical partner turn the duet into a surprise.
A coordinated gold look makes the two performers feel like one act.
01 / From portraits to performance
This workspace turns character selection into a focused video workflow. The reference clip supplies the performance, while your photos guide the two replacement identities.
Separate uploads keep the left and right references easy to follow. Check the role labels or swap the pair before starting.
The prepared clip provides the signature studio, microphone, and duet movement. You do not need to upload a video or write a prompt.
Use the original ratio or choose another supported aspect ratio. Pick 480P, 720P, or 1080P to suit your preview or download.
See the credit cost before submission, follow the generation progress, and save the completed video. Your account history keeps the task available.
02 / Three choices to make


Character references / AI concept artwork
Start with one clear photo per character. The workspace takes care of the fixed reference and replacement instructions.
Show one character per image. A clear face, good lighting, and visible outfit details help the model understand each person.
Photo 01 takes the original left role; Photo 02 takes the right. Swap them if you want the performers to trade sides.
Check the ratio, resolution, and displayed cost. Once the video is ready, inspect both characters and the final seconds before downloading.
03 / Your first duet
Understand the fixed template, photo requirements, and video settings before you generate. You can also visit Hotel Lobby AI.
The effect reimagines the orange-studio Hotel Lobby performance with a new pair of characters. This workspace uses a fixed reference clip, two separate photos, and Wan 3.0 Standard to create a 10-second take.
No. The reference performance is already included. Upload only two character photos, then choose an aspect ratio and resolution.
The first photo replaces the original performer who starts on the left; the second replaces the one on the right. The swap button reverses this assignment before you generate.
The workspace accepts JPG, PNG, WebP, and BMP images up to 20 MB each. Use a single visible person or character per photo and avoid face coverings or heavily cropped images.
The prepared instructions ask the model to preserve the orange studio, duet movements, and original sequence while replacing both people. Generated results may vary, so review the cast, framing, and ending.
Videos are fixed at 10 seconds. The available resolutions use Wan 3.0 Standard pricing: 480P costs 80 credits, 720P costs 160 credits, and 1080P costs 320 credits. Sign in with enough credits to submit.
Original ratio follows the source performance. Choose 9:16 if you need a vertical video, or another supported format for your use. A different ratio changes the framing, so check that both performers remain visible.
Yes. Download the video once generation finishes, or find the task in your account history. If you generate again with another pair or format, the new task uses the credit cost shown in the workspace.
Your next character pairing
Choose your pair and make the Hotel Lobby performance their next scene. Upload two photos to start your 10-second duet.
Create this pairingEntdecke ähnliche Tools und bleibe im kreativen Flow.
30-sekündiges audiovisuelles Storytelling mit multimodalen Referenzen und präziser Bearbeitung
Multimodales KI-Video mit nativem Audio und konsistenter Multi-Shot-Steuerung
Schnellere, kostengünstigere Seedance-Videogenerierung für schnelle Iterationen
Langformatige audiovisuelle Kreationen mit multimodalen Referenzen und präziser Bearbeitung
Unterstützt Video von erstem/letztem Frame, Multi-Bild-Referenzgenerierung
KI-Videogenerierung mit mehreren Szenen, nativem Audio und Referenzkontrolle
Entdecken Sie Kling 4.0 im KI-Video-Workspace für Shot-Planung, visuelle Referenzen, Kameraführung, Kontinuität und Ton.
Erstellen Sie mit Kling 3.0 Einzel- oder Mehrfachshots mit Start- und Endbild, wiederverwendbaren Elementen, Formatwahl und Audio.
Erstellen Sie kurze Videos mit koordinierten Bewegungen und Sound aus Text oder Bildern
