Paste a voiceover script and get one hyper-realistic image for every moment of it. Then drop in the SRT, and every image lands on a Premiere Pro timeline in order, timed to the narration.
Start with step 01 below
Step 01 · Script
Paste your script.
AI reads the whole script, understands it, and splits it into complete moments at the cut speed you choose. Each moment becomes one image.
0 words
toimages
Step 02 · Prompts
Write a prompt for every moment.
Each moment becomes one detailed, hyper-realistic prompt. Every prompt stays editable.
Split a script in step 1 and the sentences appear here.
Step 03 · Images
Generate the images.
Choose the model, frame shape and quality, then generate every image with Runware.
GPT Image uses its own fixed sizes, usually costs several times more than FLUX, and its model ID isn't confirmed yet. Press Generate on one image card first to test it and see the price.
Generated Images
Images are named in script order (001.jpg, 002.jpg…) so they stay in sequence everywhere.
Step 04 · Export
Export to Premiere Pro.
Upload the SRT and download the XML. The subtitle timings decide when each image starts. Images butt up against each other, so the timeline has no gaps.
In Premiere Pro: File, Import, pick the .xml. If you left the folder blank, Premiere will ask you to locate 001.jpg once and then finds the rest.