Fireworks
Fireworks — interactive fireworks display.
What I asked each model — the Fireworks prompt
Every model on this page got this exact prompt inside the Agent Operating System: Fireworks — interactive fireworks display.
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. 23 frontier models have attempted it so far: Claude Fable 5, Fugu Ultra, Fusion, Gemini 3.6 Flash, GLM-5.2, GPT-5.6 Sol, Grok, Inkling, Kimi K3, MiniMax M3, Hermes MoA, Opus 4.8, Claude Opus 5, Qwen 3.8, Qwen 3.7, Claude Sonnet 5, DeepSeek V4 Pro, DeepSeek V4 Flash, Fugu Mini, Kimi K2.7, Kimi K2.7 · Fast, Kimi K2.7 · No-Think, Kimi K2.7 · Quality.
Why this task matters. Fireworks 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 Fireworks
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: Clean starfield with subtle gradient and a nicely styled title, plus solid physics/palette variety in code (rings, sparks, flash), but the captured moment shows an empty sky with just one faint burst starting in the corner, so the screenshot lacks the wow-factor of a mid-explosion capture. Competent and shippable but not visually spectacular enough to top the field.
What I saw: Ultra v2 (gap-fill) — click-to-launch fireworks. Smoke-test PASS with 26.3% pixel diff — among the most reactive builds.
What I saw: Click to launch fireworks — particle trails, sparkle physics, AND a synthesized whoosh + boom via Web Audio. Most polished fireworks of any model.
What I saw: Strong ambitious 3D Three.js setup with city silhouette, star field, polished UI, shell-style selector, and synthesized audio — but the screenshot only captures faint rising rocket trails with no actual bursting fireworks visible, so the core payoff isn't demonstrated on screen. Good polish and interactivity, but the empty-looking sky keeps it from the top tier.
What I saw: Gorgeous rendered scene with a gradient night sky, twinkling stars, lit city skyline, multicolored rocket trails and detailed bursts plus polished title/hint UI — clearly a top-tier, on-brief interactive build. Only minor nit: the barrage of simultaneous rocket lines looks slightly busy in the static frame, but the craft and depth are outstanding.
What I saw: Click-to-launch fireworks with particle trails, sparkle physics. 14KB. No audio (vs Fusion's whoosh+boom).
What I saw: Polished neon Orbitron title and clean glassmorphic hint render well, and the Three.js particle/explosion system with raycasted click-launch and auto-spawn is solid; but the screenshot shows an empty sky with no visible fireworks yet, so it reads generic and unproven versus the field's best animated captures.
What I saw: Gorgeous rendered scene with a detailed city skyline, moon, stars, and a rich variety of firework shapes (rings, willows, flowers, hearts, chrysanthemums) all glowing convincingly; the polished HUD and multiple simultaneous bursts make it a task winner, with only slight visual noise from overlapping trails as a minor nit.
What I saw: Click-to-launch fireworks with particle trails.
The winner on Fireworks
Fugu Ultra took gold on this task. winner · fireworks.
What I saw: Ultra v2 (gap-fill) — click-to-launch fireworks. Smoke-test PASS with 26.3% pixel diff — among the most reactive builds.
See Fugu Ultra's full model card: /models/fugu.
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
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVE
▶ LIVEHow I scored Fireworks — 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 Fireworks · 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 4,000+ founders shipping with it every day all live inside the AI Profit Boardroom.