Studio renders

Studio turns views of your scene into photorealistic images and short video clips. You capture a snapshot of your model, describe the look you want, pick a model, and generate — the AI keeps your real layout while adding materials, lighting, and atmosphere.

​

Switch to Studio

Open a project and use the Edit / Studio toggle in the navbar. Studio replaces the editing rail with a Generate panel and swaps the bottom toolbar for a capture bar. Your credit balance shows next to the Studio toggle.

​

Capture a snapshot

Snapshots are how the AI sees your actual scene. Frame the view you want in the viewport, then use the capture bar at the bottom:

• Standard — a fixed aspect ratio (16:9, 9:16, 4:3, 3:4, or 1:1) at high resolution. Click Standard again to switch aspect.
• Viewport — exactly what your canvas shows, at canvas size.
• Area — drag to capture just part of the view.

Press Capture to enter capture mode with a framing overlay, then take the shot. The snapshot lands in the Generate panel as a reference.

​

Compose a generation

The Generate panel is where you set up the render:

• Image or Video — the switch at the top picks the output type.
• Prompt — describe the result. The prompt is smart: highlighted modifiers can be clicked to swap their value, and the Enhance button rewrites your draft into a stronger prompt.
• Model — the model card opens a picker of image and video models, each with a short description of its strengths, a speed estimate, and its credit price. Some models are better at preserving your exact layout; others at restyling or motion.

​

References

References steer the result. Use Add reference to take a snapshot, pick a past render from the gallery, or upload an image. You can give each reference a role:

• Structure — locks your real layout and proportions (typically your snapshot).
• Style — carries look, color, and materials (a render or photo you like).
• Composition — guides framing and element placement.

Video models take input differently depending on the model: frame-based models animate from a start frame (optionally toward an end frame), while reference-based models compose the shot from a set of reference images.

​

Output settings

Depending on the mode and model you’ll also see:

• Variations — generate 1 to 4 images per run.
• Quality or resolution — output tiers on models that support them; higher tiers cost more credits.
• Aspect ratio — when the model can honor one.
• Motion length and resolution — for video, the clip duration and output resolution.

​

Generate and track

The generate button shows the credit cost of the run before you commit — count (or seconds) × the model’s per-unit price, next to your remaining balance. While a render is in flight it appears in the in-progress list at the bottom of the panel with a rough time estimate; you can keep composing the next one.

​

The gallery

The Scene / Gallery switch at the top of the stage flips the viewport between your 3D scene and everything you’ve generated in this project. Open a render to:

• Download it.
• Use as reference — feed it back into the next generation to iterate on a look.
• Add to chat — send it to the AI assistant to discuss or act on.
• See the prompt and references that produced it.

​

How credits work

Generations are metered in credits:

• Each model has its own price, per image for image models and per second for video models — it’s shown on the model card and in the cost line under the generate button. Higher quality and resolution tiers cost more.
• You get free credits every day. The free allowance tops back up daily at midnight UTC and doesn’t roll over.
• Purchased credits never expire and are only spent after your daily free credits run out. Buy packs from the in-app credits screen — checkout supports promotion codes.
• If a run costs more than you have left, the generate button becomes Buy more credits.

See Credits for balances, packs, and purchase history.