Command Code Tests Mimo-V2.6-Flash, Grok-4.7, and DeepSeek-V4.1-Flash on 3D SubwayBench

Key Info

Command Code's engineering team compared Mimo-V2.6-Flash, Grok-4.7, and DeepSeek-V4.1-Flash on the 3D SubwayBench benchmark using the same /design prompt. DeepSeek-V4.1-Flash scored 9/10 at $0.024 with strong one-shot 3D UI/UX and gameplay, while Grok 4.7 scored only 3/10 at $0.35 needing more iterations.

Highlights

• DeepSeek-V4.1-Flash: 9/10, $0.024, one-shot success, strong 3D UI/UX and gameplay.
• Mimo-V2.6-Flash: 8/10, $0.018, playable 2D result in ~4 iterations.
• Grok 4.7: 3/10, $0.35, slower UI and gameplay, requires more iteration.
• Team has run 26+ side-by-side comparisons across frontier and open models; all runs are public and open source.