Image
Image turns a written description into a finished image and drops it straight into your composition as a layer. Reach for it when you need a background, a textured plate, an isolated subject, or a quick concept frame without leaving After Effects or hunting for stock. You describe what you want, and Prism generates the image and places it on the timeline at your playhead, scaled to fit your comp.
How to use it
Make sure a comp is open and active, and move the playhead to the moment where you want the image to appear. The image is placed at the current time.
In the MCP panel, switch to the Prism AI tab and find Image in the list.
Run Image and a prompt window appears. Describe the subject, style, and lighting: for example, "neon-lit rainy Tokyo alley, cinematic, 16:9." If you want an isolated subject with no background, ask for a transparent background in your description.
Prism generates the image and imports it as a layer at the playhead, centered and scaled to fit your comp. The whole step is grouped as a single undo, so one undo removes it cleanly.
You can also just ask your AI to do it for you. With Prism connected, tell your AI something like "generate a foggy forest background and drop it into my comp," and it runs Image on your behalf and places the result the same way.
What you get
- A new footage layer in your active comp. The generated image is imported and added as its own layer, named from a short version of your prompt. It never touches or replaces any existing layer.
- Placed at the playhead. The layer's start time matches the current time, so it begins exactly where your playhead sits.
- Centered and scaled to fit. The image is positioned in the middle of the frame and scaled to fit inside your comp without cropping. Larger or smaller comps are handled automatically.
- Optional transparent output. Ask for a transparent background in your prompt to get an isolated subject with a real alpha channel: useful for overlays, cutouts, and composites. By default, images come back as solid, opaque frames.
- Aspect ratio in the prompt. There is no separate ratio control: phrase it in your description (for example, "16:9" or "square") to steer the framing.
A composition must be open before you run Image; if no comp is active, Prism asks you to open one first. The generated file is written to a temporary location and imported, so the layer references that file like any other footage.
Notes and limits
- One image is generated per run.
- Generation happens on Prism's servers, so the tool needs your panel connected. If it can't reach the server, it tells you to check the connection. See troubleshooting.
- Image has no fixed rate: it depends on the model your AI client chose, so a small model costs a fraction of a large one. The exact amount for every generation appears in your dashboard activity. See usage and limits for how the balance is metered across every tool.
Related
- Usage and limits: plan caps, how metering works, and where to see the current rates
- Subtitles from audio: another balance-metered Tools-tab capability
- The Prism AI: where Image lives in the panel
- Plans and pricing: Free vs Pro and what each includes