Renderly generates a video's images from prompts in one call

Renderly now generates the images in a video from text prompts inside the render call, in one API request. The company says the feature, AI images, is live today and turns a stack of prompts into the finished video without leaving its API.

Renderly announced it in a blog post dated October 18, 2026, written by founder Phil Duong. An image variable in a template now accepts {"image": {"prompt": "..."}} where it used to take an image URL. Renderly generates the images, places them in the template and renders the video. The post calls it "one API call, prompts in, finished video out."

The controls are small. A top-level imageStyle puts one look on every image in a render, and a per-image style field overrides it for a single slide. In the post's demo, slide four carried its own style and came out as a watercolor while the other five kept the shared photo look. One render takes up to 20 image prompts, and Renderly says they can be mixed freely with text, media, voiceover and caption values.

The demo used the public "Destination Guide - Top 5 Things To Do" template, which has six image slots. Renderly says the request returned in 39 seconds and the finished 34-second video was charged 3.7 credits in total: 2.7 for the images and 1 for the render. A repeat of the same request, the post says, cost 1 credit instead of 3.7 and finished in 11 seconds, because an identical request returns the stored image for free.

Credits are Renderly's own unit, drawn from the same balance as rendering, according to its AI features guide. The post and the guide give per-image credit costs but no dollar value for a credit: 0.5 credits per quality image, 0.2 per fast image, double for 2K, charged only when an image succeeds.

Renderly is a programmatic video API that renders videos at scale from JSON-configurable templates. Its rendering guide lists AI replacement values as a way to generate assets inside a render, and the AI features guide counts AI images as the third of three AI building blocks in the pipeline, after auto-captions and AI voiceover. Alongside the in-render path there is a standalone endpoint, POST /api/v1/ai/images, that takes a prompt and returns one image URL saved to the media library, and an MCP tool named renderly_generate_image for Claude and other assistants, available on API-key connections today.

Duong's stated reason for building it: users were generating images in one tool, hosting them, then sending the URLs to Renderly to make a slideshow. The post describes that as "two products, two bills and two integrations for one video."

For a creator, the change could remove a hosting step and a second bill from a slideshow workflow, if the images hold up in a finished render. Renderly's own figures cover cost and time, not image quality, and the post shows six images from a single template.

AI images are live today. API keys are self-serve from the Renderly dashboard under Settings, API Keys, and new accounts start with free credits to test with.