Plasma
Plasma — hypnotic full-screen plasma effect with palette switcher, click ripples.
What I asked each model — the Plasma prompt
Every model on this page got this exact prompt inside the Agent Operating System: Plasma — hypnotic full-screen plasma effect with palette switcher, click ripples.
Single HTML file out. No iteration. No examples in the system prompt. Whatever each model produced on the first run is what's on this page. 6 frontier models have attempted it so far: Fusion, Gemma-4 12B Coder, Grok, Kimi K2.7, MiniMax M3, GLM-5.2.
Why this task matters. Plasma is a textbook test of visual-class capability — the kind of build that exposes whether a model is doing pattern-matching or actual reasoning. Shipping this cleanly is the floor for what I expect from a frontier model — every model on the leaderboard should at least attempt it.
How each model handled Plasma
Ranked by my 0–10 score from the source comparison guides on agentos.guide. Click any to play the actual one-shot HTML the model produced.
What I saw: Hypnotic full-screen plasma effect with 5 palettes, click anywhere to create a ripple. Smooth slow motion. 13KB but the brief is fully met.
What I saw: Broken. It gets a WebGL context and writes vertex/fragment shader source into <script> tags, but never compiles, links or draws them (no createShader, no drawArrays). Nothing renders.
What I saw: Full-screen plasma with palette switcher + click-to-ripple. Lighter build (9KB) than Fusion's.
What I saw: Plasma effect with palette switcher + click ripple. Cleaner than M3's stub.
What I saw: Minimal 1KB plasma. Brief is barely met.
Demo on the bench. Not scored yet — play it and form your own opinion.
The winner on Plasma
Fusion took gold on this task.
What I saw: Hypnotic full-screen plasma effect with 5 palettes, click anywhere to create a ripple. Smooth slow motion. 13KB but the brief is fully met.
See Fusion's full model card: /models/fusion. Direct head-to-head against the runner-up: Fusion vs Grok.
Every attempt — live, playable
Side by side. Click any tile to run that model's actual one-shot HTML in a new tab.
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVEHow I scored Plasma — methodology
Three axes, 0–10 each, averaged. Runs: drop the .html in a browser; if it opens to a broken page, it scores zero. Hits the brief: did the model ship the thing the prompt asked for, or a different thing it found easier. Looks good: visual polish, motion, interactivity — where most of the gap between gold and silver lives.
My scores trace back to the source comparison guides on agentos.guide. See the full methodology page for data provenance, including which source guide each cell's score came from.
Related
More visual benchmarks: all tasks in the Visual category · See the best AI model for Plasma · Back to the leaderboard
Run this stack yourself.
Every demo on this bench was built inside the Agent Operating System — one prompt, one shot, single HTML file out. The Agent OS, the prompts, the templates, the weekly walkthroughs and 3,600+ founders shipping with it every day all live inside the AI Profit Boardroom.