Visual / Systems Simulation

City Water System Simulator

The water-system prompt asks for a city dashboard with a reservoir, treatment plants, neighborhoods, pipe flows, and failures. When a pipe bursts or demand spikes, pressure and supply should change across the city.

Prompt

Build a 2D simulation of a city's drinking water system: 1 reservoir, 3 treatment plants, a network of pipes, and 20 neighborhoods with varying demand. A demand producer makes households and businesses use water throughout the day (morning shower peak, midday for businesses), and a distribution consumer balances flow from plants to keep pressure stable everywhere....

Max tokens
100K
temperature
0
top_p
1
seed
42
presence_penalty
0
frequency_penalty
0
Reasoning effort
High
Execution
Single-shot via API

Fortytwo Prime

Fortytwo

PASS5 / 5

Fortytwo Prime passes all five City Water System criteria, showing the full reservoir-plant-neighborhood map, pressure and flow visuals, time-based demand, event triggers, and live event feedback.

vs
Qwen 3.7 PlusAlibaba
OpenRouter Fusion (Quality)OpenRouterPASSClaude Fable 5AnthropicPASSClaude Opus 4.8AnthropicPASSChatGPT 5.5OpenAIPASSGrok 4.3xAIPASSClaude Opus 4.7AnthropicPASSGLM 5.1ZhipuPASSChatGPT 5.4OpenAIPASSKimi K2.7 CodeMoonshotMIXEDGemini 3.1 ProGoogleMIXEDDeepSeek V4 FlashDeepSeekMIXEDDeepSeek V4 ProDeepSeekMIXEDGemma 4GoogleMIXEDKimi K2.5MoonshotMIXEDGrok 4.20xAIMIXEDDeepSeek V3.2DeepSeekMIXEDQwen 3.7 PlusAlibabaMIXEDGemini 3.5 FlashGoogleMIXEDQwen 3.6 PlusAlibabaMIXEDClaude Opus 4.6AnthropicFAILNemotron 3 UltraNVIDIAFAILMiniMax M2.7MiniMaxFAIL
MIXED1 / 5

Qwen 3.7 Plus renders a pipe network with three treatment plants, neighborhood gauges, flow animation, and the required heatwave, pipe-burst, and plant-offline buttons. However, the reservoir is not clearly represented as a map element or level indicator, the system starts around 2 PSI despite full reservoir and online plants, and event effects are shallow: demand and average pressure change, but the map/telemetry do not clearly show a coherent plant-offline or reservoir response.

Model verdicts

ModelVerdictCreate Map To SpecAccurate Parameter VisualizationDemand Simulation + System ResponseEvent Triggers And EffectsReal-Time Response To Events
1Fortytwo PrimePASSpasspasspasspasspass2OpenRouter Fusion (Quality)PASSpasspasspasspasspass3Claude Fable 5PASSpasspasspasspasspass4Claude Opus 4.8PASSpasspasspasspasspass5ChatGPT 5.5PASSpasspasspasspasspass6Grok 4.3PASSpasspasspasspasspass7Claude Opus 4.7PASSpasspasspasspasspass8GLM 5.1PASSpasspasspasspasspass9ChatGPT 5.4PASSpasspasspasspasspass10Kimi K2.7 CodeMIXEDpassmixedpasspasspass11Gemini 3.1 ProMIXEDpassmixedpasspasspass12DeepSeek V4 ProMIXEDpassmixedfailpasspass13DeepSeek V4 FlashMIXEDmixedmixedpasspassmixed14Gemma 4MIXEDmixedfailpasspasspass15Kimi K2.5MIXEDpassmixedpasspassfail16Grok 4.20MIXEDmixedfailpassmixedpass17DeepSeek V3.2MIXEDpassmixedfailpassmixed18Qwen 3.7 PlusMIXEDmixedmixedfailpassmixed19Gemini 3.5 FlashMIXEDmixedmixedfailpassmixed20Qwen 3.6 PlusMIXEDpasspassfailfailfail21Claude Opus 4.6FAILfailmixedfailmixedfail22Nemotron 3 UltraFAILfailfailfailfailfail23MiniMax M2.7FAILfailfailfailfailfail

Have a complex task to evaluate?

Request a custom evaluation for your use case.

Request a demo →