Brief 71
44 shoot types across fashion, e-commerce, beauty, product, food, brand, spaces and entertainment, plus deliverable and mood.
A virtual photo studio for AI images and video.
Pick talent, set, light and lens with a click. Cycolorama writes the prompt and renders it with Gemini or OpenAI.
743 studio options
44 shoot types
44 one-click recipes
2 engine families, photo and video
From brief to render in four moves
You direct the shoot the way you would on a real set. Cycolorama turns each decision into precise prompt language behind the scenes.
-
Start from a recipe
Load one of 44 proven setups, from marketplace packshots and backlit glass to 360 product spins, or start from a blank set.
-
Give references a job
Drop in photos or a video and tag each one as subject, product, outfit, pose, style, location or layout. Or skip references entirely.
-
Direct every department
Cast the talent, dress the set, place the lights, choose the lens and set the grade. Every click shows up in the shot sheet.
-
Press the shutter
Review the generated prompt, add extra direction if you like, and render. Results land on your contact sheet, saved on your server.
Everything a real set has, one click away
Nine departments, each written in the language photographers and directors actually use, so the model gets instructions it understands.
Talent 145
Cast size, adult age ranges, 22 heritages, 9 skin tones, body types, hair, makeup, 21 poses, expression and gaze.
Styling 92
Tailoring, streetwear, evening, activewear and traditional dress from saree to kente, with fabrics, accessories and color stories.
Props 101
Furniture, plinths, botanicals, textiles, lifestyle and tech props, plus practical effects like haze, splash, petals and levitation.
Set 93
Seamless paper in 14 colors, cycoloramas, textured walls, interiors, urban and natural locations, floors, time of day and weather.
Lighting 81
Rembrandt, butterfly, clamshell, chiaroscuro and window light, with modifiers, rim and hair lights, gels, Kelvin and ratios.
Camera 73
Digital, cinema and film formats, 14 lenses from fisheye to anamorphic, aperture, framing, angle, composition and technique.
Grade 40
Film stock looks like Portra and CineStill, color grades, grain and halation, and retouching from natural to product perfect.
Motion 47
For video: push-ins, orbits, crane and drone moves, subject action, slow motion and timelapse, sound and pacing.
Built like a pro tool, not a chat box
Bring your own engine
Connect Gemini, OpenAI or any compatible endpoint. Switch engines from the top bar at any time.
Photo and video in one place
Flip to video and the same shot gains camera moves, action, speed and sound.
Works on any screen
Direct a shoot from a desktop, a tablet on set or your phone.
Read references
A vision model studies your reference frames, writes a shot breakdown and presses the matching set, lighting, camera and grade buttons for you. It never guesses identity traits; you choose those.
Your options, your recipes
Add your own buttons to any category and save looks you like as recipes for one-click reuse across a whole campaign.
Works with the models you already use
- Google Gemini
- Nano Banana image models and Imagen for stills, with reference images attached. Veo 3.1 for video from a starting frame or asset references.
- OpenAI and compatible APIs
- gpt-image for generation and multi-reference edits with high input fidelity. Local OpenAI-compatible servers work too.
- New models, no update needed
- Model names are free text with suggestions, and a connection test lists every model your key can use.
Private and secure by default
- API keys are encrypted with AES-256-GCM on your server and never reach the browser.
- Personal accounts with email verification. Each user's keys, renders and library are stored separately.
- Location and camera metadata are stripped from photos before upload.
- Protection against cross-site requests, internal network access and oversized uploads.
- Your private dashboard shows every render, its cost driver and any key errors, and you can sign out other devices at any time.
- Generated people are adults, and real-person references require confirmed consent.
Questions
Which AI models does Cycolorama work with?
Google Gemini image models (including Nano Banana and Imagen), Veo for video, and any OpenAI-compatible API, such as gpt-image for images. You can also point it at a local OpenAI-compatible server.
Do I need to write prompts?
No. Every button adds a precise instruction, and Cycolorama assembles a structured prompt in the background. An optional direction box covers anything the buttons don't, and the full prompt stays visible and editable.
Can I use my own reference photos or video?
Yes. Upload images or a short video and tag each as subject, product, outfit, pose, style, location or layout. Videos are split into keyframes. Photo metadata is stripped before upload.
Where are my API keys stored?
On our server, encrypted with AES-256-GCM using a key unique to your account. Keys are only decrypted at the moment a render is sent, and never sent back to the browser.
How much does Cycolorama cost?
Cycolorama is free during early access. You connect your own Gemini or OpenAI API key, the provider bills you directly at its own rates for what you render, and Cycolorama adds nothing on top.
Does it work for any industry?
It ships with 44 shoot types across fashion, e-commerce, beauty, product, food and drink, corporate and brand, automotive, real estate, hospitality and entertainment, and you can add your own options to any category.
How do I send feedback or request a feature?
Email partnerships@zesterproductstudio.com or message @cycolorama on Instagram. Tell us what got in the way and which features, recipes or models you want next. Cycolorama is in early access and we read every message.