3D pixel-art streetball court comparison
- Views
- 8.4K
- Likes
- 154
- Replies
- 29
- Saves
- 80
A 3D pixel-art isometric streetball courtyard simulation built with three.js. Claude Opus 5.5 and GPT-6.1 Sol were compared on the same prompt, with Sol given one revision.
Full prompt2,643 chars
Build a real-time pixel-art isometric diorama of 3x3 streetball at a single hoop in a city courtyard, as a single HTML page using three.js (WebGL). A floating island of asphalt holds a painted half court (faded blue with an orange key, worn white lines, the two-point arc) inside a box of low green boards. One hoop stands on a steel pole with a steel backboard, an orange ring and a chain net, with a wire fence behind it. Around the court: a three-storey brick block of flats with balconies and washing, two grannies on a bench by the door, pigeons on the roof. A row of tin garages with original graffiti tags, and an old boxy car half out with its bonnet up and a man working on it. A playground with working swings, a slide and a sandpit, plus a mother on a bench. A kvass barrel with a seller, poplars, floodlight masts, and a wire with a pair of sneakers hanging from it. Render at low resolution with an orthographic camera, snap colours to a fixed 64-colour OKLab palette and draw depth outlines. After rain, puddles reflect the scene. Simulate the ball physically: drag, lively bounces off dry asphalt and duller ones off wet, a steel backboard, the ring as a real torus (rim-outs, rolls, lips) and a chain net that swishes. Solve every jump shot numerically through the flight with drag, at an arc that steepens close in, then add hand error that grows with distance, a defender's contest and a weaker shooter. Make layups steep; put dunks through from above with the shooter hanging on the rim; throw alley-oops as lobs to a point over the ring. AI: dribbling with a hand that rides the ball, crossovers, drives, open-shot decisions, passes (chest, bounce, lob), cuts and screens. Man-to-man defence with help on drives, reach-in steals, contests and blocks, box-outs and rebounds caught with a jump. Rules: ones inside the arc, twos outside; after a make the other team takes the ball from under the ring and must clear it past the arc; check at the top after out-of-bounds. First to 11 or 21. Three teams: winners stay, the losers walk off to the bench and the next team walks on. Life in the yard: grannies feed the pigeons and complain about the boombox, the waiting team nods to the music and goes for kvass, a cat stalks the pigeons, children swing and slide. Day/sunset/night with floodlights and lit windows; dry, after-rain puddles, or rain with rings. UI: a scoreboard; a side rail with three team ladders, called plays (dunk, deep two, alley-oop), weather and a boombox toggle, field-goal stats and a live feed; click the court to mark the next pass. Original characters only, English UI, responsive down to phone width.
Source: the creator's reply under the post. Kept in its original language, unedited.
Want more prompts like this one?
Get the 50 most-viewed Claude 5.5 video prompts on one page: the full original text, the creator and the post it came from. We'll only email you when there's a new prompt worth trying. Unsubscribe anytime.
Tools the creator mentioned
Original post: https://x.com/vib3coded/status/2115106186831823065
All works are open data at GitHub · awesome-opus-5.5-video.
Model attribution is as stated by the creator and was not independently reproduced. View counts are a snapshot from 2026-10-29. The work belongs to its creator; DM @GoSailGlobal on X for corrections or removal.
More comparisons
0:42AI self-portrait challenge: Sonnet 5.5 vs GPT-6 Astra
7.3M views
0:28Opus 5.5 vs GPT-6 Astra samurai game demo
1.1M views
0:13Gummy bear soft-body physics scene
630.4K views
0:30GPT-6 Astra vs Claude Opus 5.5 comparison
604.6K views
0:22Grok vs Opus grass-touching simulator cost test
554.3K views
0:44Opus 5.5 vs GPT-6 Sol Unreal game dev
543.6K views
Related articles
OpenAI Academy vs Anthropic: A Complete Comparison of Two Leading AI Companies' Free Learning Paths
Both of the top AI companies offer official free learning platforms — together that's 25+ courses and hundreds of hours of content, enough to take you from AI newcomer to resume-worthy engineer. This post maps out all 15 OpenAI Academy courses along a 10-week roadmap and provides a full side-by-side comparison with the Anthropic learning path.
OpenAI Agents SDK Major Upgrade: Harness and Sandbox Now Built Into the SDK — The Opposite Approach from Anthropic
OpenAI has shipped a major upgrade to its Agents SDK, with the headline changes being harness and sandbox support built directly into the SDK. Anthropic's Claude Agent SDK takes exactly the opposite approach — one gives you everything pre-assembled, the other hands you a toolbox.
Three Companies Worth Watching for Cutting-Edge Agent Development
If you had to pick three companies to follow for the latest frontiers in AI agents — Anthropic, OpenAI, and Manus — each occupies a distinct position.
Andrew Ng's 2026 AI Prompting Course – Module 3 Complete Notes: Let AI Generate Images, Build Apps, and Analyze Data
Complete notes for Andrew Ng's AI Prompting for Everyone, Module 3. Six lessons covering multimodal cost awareness, image understanding and generation, building mini apps with prompts, AI-assisted coding and data analysis, and a final project walkthrough. Includes 9 ready-to-use prompt templates.