Gtafoot
GTA On-Foot — third-person city streets on foot: weapons, shooting, pedestrians, cover, wanted stars.
What I asked each model — the Gtafoot prompt
Every model on this page got this exact prompt inside the Agent Operating System: GTA On-Foot — third-person city streets on foot: weapons, shooting, pedestrians, cover, wanted stars.
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. 25 frontier models have attempted it so far: Claude Fable 5, Fugu Ultra 1.1, Fusion, Gemini 3.6 Flash, GLM-5.2, GPT-5.6 Sol, Grok, Grok 4.6, Hy3, Inkling, Kimi K2.7, Kimi K3, MiniMax M3, Hermes MoA, Muse Spark 1.2, Opus 4.8, Claude Opus 5, Qwen 3.8, Qwen 3.7, Claude Sonnet 5, DeepSeek V4 Pro, DeepSeek V4 Flash, Kimi K2.7 · Fast, Kimi K2.7 · No-Think, Kimi K2.7 · Quality.
Why this task matters. Gtafoot is a textbook test of game-class capability — the kind of build that exposes whether a model is doing pattern-matching or actual reasoning. A model that ships this in one shot is usually safe to wire into your agent loop for harder tasks of the same shape.
How each model handled Gtafoot
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: One-shot: night city w/ dense lit windows, DINER neon, rain streaks, lamp pools, blocky humanoid player, cohesive HUD + minimap + wanted stars. Dark at street level but atmospheric. Eyeballed.
What I saw: Polished HUD (health/ammo/wanted stars/minimap with tracked entities) and a well-modeled character with active shooting (ammo already at 089, 'CIVILIANS SCATTER' banner), but the camera is clipped hard into a building wall showing mostly empty geometry and no visible enemies/combat in frame, so it reads more like a functional walking sim than the shootout the brief demands.
What I saw: SHOWCASE BUILD (threejs-game-director): articulated humanoid w/ walk cycle + weapon poses, 110 instanced buildings w/ lit windows, 16 neon blade signs, wet-road reflections, cops+peds AI, cohesive GTA HUD (cash/minimap/stars/weapon wheel), 71fps. Eyeball-gate passed.
What I saw: Strong atmospheric dusk city with polished HUD (health/stamina/ammo/kills, wanted stars, minimap), visible pedestrians, working cover system, and crosshair — but the screenshot shows no active combat or enemies engaging, kills at 0, and it reads more as a walking sim with cover than a lively shooting sandbox.
What I saw: 43KB · plays clean · plain (re-rolled)
What I saw: The screenshot shows a broken top-down camera view (not the intended third-person on-foot perspective) — the world renders as a flat blurry grid with no visible 3D buildings, pedestrians, or player character, only a stray triangle marker; HUD elements (health, weapon, ammo, hint) are present but the core rendered scene fails the brief. Strong HUD/code ambition, but the actual render is essentially non-functional as a GTA on-foot experience.
What I saw: 10KB · plays clean · three, webgl, input (re-rolled)
What I saw: Third-person street level with a real armed character model, long cast shadows, a sunset skybox, pedestrians and parked cars down the block, plus a grid minimap. Played 12s: AMMO went 11 to 09 and SCORE 000010 to 000030 on click-fire, so shooting and scoring are wired. World is blocky and the buildings are flat-faced.
What I saw: Strong dusk-city atmosphere with a readable blocky hero (cap, jacket trim, gun), streetlights, crosswalks, and a pedestrian, plus clean HUD (ammo, health bar, wanted stars, crosshair, controls). Weak points: the world is fairly barren mid-frame, no visible buildings-as-cover in this angle, and it reads more as a stylized diorama than the dense GTA street the brief implies, keeping it below the field's best.
What I saw: Renders a clean neon 3D city with roads, cover barriers, pedestrians and a wanted-star HUD, but the giant title/subtitle/DOM cover-barrier overlay dominates the frame and the scene reads generic voxel blocks rather than a polished GTA-style on-foot game; functional but not a task winner.
The winner on Gtafoot
Fusion took gold on this task. showcase · game-director.
What I saw: SHOWCASE BUILD (threejs-game-director): articulated humanoid w/ walk cycle + weapon poses, 110 instanced buildings w/ lit windows, 16 neon blade signs, wet-road reflections, cops+peds AI, cohesive GTA HUD (cash/minimap/stars/weapon wheel), 71fps. Eyeball-gate passed.
See Fusion's full model card: /models/fusion. Direct head-to-head against the runner-up: Fusion vs Claude Opus 5.
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
▶ LIVE
▶ LIVE
▶ LIVEHow I scored Gtafoot — 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 game benchmarks: all tasks in the Game category · See the best AI model for Gtafoot · 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.