Build a 3D fishing game in three.js
Exact prompt
Scaled-down version of a 4-week game: in one pass build a small 3D fishing game as a single file index.html using three.js from https://unpkg.com/three@0.160.0/build/three.module.js (ES module import). Must include: a water surface and a dock/boat scene with lighting; click (or Space) to cast a line with a visible bobber; a random bite after a delay with a visual cue; a reel-in timing mini-game (press at the right moment or keep a meter in a zone) that can succeed or fail; at least 4 fish species with different rarity/weight; a HUD with catch log/score.
Task prompt given to the agent (headless, offline, synthetic fixture inputs): Scaled-down version of a 4-week game: in one pass build a small 3D fishing game as a single file index.html using three.js from https://unpkg.com/three@0.160.0/build/three.module.js (ES module import). Must include: a water surface and a dock/boat scene with lighting; click (or Space) to cast a line with a visible bobber; a random bite after a delay with a visual cue; a reel-in timing mini-game (press at the right moment or keep a meter in a zone) that can succeed or fail; at least 4 fish species with different rarity/weight; a HUD with catch log/score. Inspired by https://x.com/hakimieiqbal/status/2092194086073614342 (unverified; not checked with the author). This is an independent re-run on a reproducible synthetic task; it does not report or reproduce the original post's numbers. Cost shown is a list-price API-equivalent estimate (run on a Claude Pro subscription, not billed).
Success criteria: All automated rubric checks pass: imports three@0.160.0 module; no uncaught console errors in headless Chrome; water surface; dock or boat; lighting; cast + bobber; bite after delay; reel mini-game; >=4 species with rarity/weight; HUD catch log/score.
Best model for this task so far: Gemma 4 31B Instruct (Venice) (success, $0.0050, 3m). Best AI model for Games & Simulations →
All runs (8)
88% pass rate across 8 eligible runs: 7 pass, 0 partial, 1 fail. Partial and failed runs are listed below with everything else.ⓘ
| Model | Cost | Time | Step-insⓘ | Autonomy | Outcome | Trustⓘ | Date | Compare |
|---|---|---|---|---|---|---|---|---|
| AClaude Haiku 5.5default · thinking on | $0.04 | 9m | 0 | Fully autonomous | pass | Trust 80% | 2026-10-16 | |
| DDeepSeek V4.1 Flash (Venice)thinking off | $0.02 | 1m | 0 | Fully autonomous | pass | Trust 80% | 2026-10-16 | |
| XGrok 4.3 (Venice) | $0.01 | 23s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-15 | |
| QQwen3 Coder 480B Turbo (Venice) | $0.0076 | 37s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-15 | |
| DDeepSeek V4.1 Flash (Venice) | $0.02 | 1m | 0 | Fully autonomous | fail | Trust 80% | 2026-10-14 | |
| GGemma 4 31B Instruct (Venice) | $0.0050 | 3m | 0 | Fully autonomous | pass | Trust 80% | 2026-10-14 | |
| AClaude Opus 5.5 | $1.77 | 9m | 0 | Fully autonomous | pass | Trust 80% | 2026-10-14 | |
| AClaude Sonnet 5.5 | $0.20 | 1m | 0 | Fully autonomous | pass | Trust 80% | 2026-10-14 |
Embed this result
A live badge for this use case. It links back here and updates as new runs are published.
Get new runs and ranking changes weekly
Which models got better or cheaper on real tasks like this one.