MODEL GAUNTLET / COMPETITOR PROFILE
MiniMax M3.1 Flash.
Rank 3 of 4 · 424.9/500 across B01–B05 · Wildcards 150/200.
B01 / Mobile OS
All models on B01 ↗TECHNICAL
89/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · FUNCTIONALITY
96/100Band: Good
FINAL
91.8/100Not capped
MiniMax M3.1 Flash delivers a fully self-contained mobile OS simulation (Vela OS) with 6,351 LOC across 22 files, implementing all 8 core domain features (lock screen, phone, messages, contacts, notes, calculator, gallery, settings) with a shared persistence layer and cross-app integration. All 22 automated browser-driven QA checks pass, 39 screenshots provide visual evidence, and zero external dependencies or eval() usage confirm clean hygiene.
- Time
- 49 min
- Tokens
- 31,156,799
- Client / effort
- MiniMax MCode
- Finished
- 22 October 2026
- Brief 1.0 · prompt SHA-256
- a00e95f7a9243ba2b2a9046d9ff7225c565196138c18322f4a0c659241bf9755
B02 / Blender Mining Drone
All models on B02 ↗
TECHNICAL
95/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · VISUAL QUALITY
82.5/100Band: Good
FINAL
90/100Not capped
Strong submission with 149 distinct editable meshes, 7/7 required subsystems detected, 22 PBR materials, and 7 rendered views. Main deficiencies: hazard markings not verified in palette analysis, glTF 2.0 compliance flagged false by automated check (though QA claims Godot import passed), and 148/149 meshes missing UVs limiting texture application in downstream engines.
- Time
- 1 h 2 min
- Tokens
- 52,775,072
- Client / effort
- MiniMax MCode
- Finished
- 22 October 2026
- Brief 1.0 · prompt SHA-256
- 4d9e8edee96c488962673f7c21868a8cba3ea55012bbef6c9e733f505dd413b3
B03 / Blender to Godot
All models on B03 ↗TECHNICAL
97/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · PLAYABILITY
84/100Band: Good
FINAL
91.8/100Not capped
Exceptional submission with exact B02 GLB provenance (SHA-256 verified), 149-mesh/22-material drone imported into a well-structured Godot 4.7 project with 14 gameplay scripts (4994 LOC), 12 scenes, 8 audio files, and 31/31 automated QA checks passing. All critical requirements met: project launches cleanly, correct drone imports, movement/collision/mining/win-state logic fully implemented with mechanical animations and audio feedback.
- Time
- 1 h 5 min
- Tokens
- 51,862,619
- Client / effort
- MiniMax MCode
- Finished
- 22 October 2026
- Brief 1.0 · prompt SHA-256
- f00af20d56033bae25fc2d36a139ec5f56e08b50d99be07946ad19c06a5f5527
B04 / Prime London Estate Agent
All models on B04 ↗TECHNICAL
85.5/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · DESIGN & NAVIGATION
50/100Band: Acceptable
FINAL
71.3/100Not capped
Wren & Sloane is a well-engineered static estate agency site with 137 files, 8555 LOC, bundled open-licensed fonts, responsive WebP images at multiple breakpoints, and documented test suites (86 unit + 25 e2e). The codebase demonstrates strong architecture with dedicated modules for filters, shortlist, and validation. The primary limitation for scoring is the absence of screenshots, preventing full visual verification of brand consistency, layout polish, and responsive rendering.
- Time
- 1 h 29 min
- Tokens
- 82,393,005
- Client / effort
- MiniMax MCode
- Finished
- 23 October 2026
- Brief 2.0 · prompt SHA-256
- b2229dc93c7a8d4ec7dda8712ff98e2f0c8cbc618181cd8cc2239ac458d1bc50
B05 / Oliver GPU Game
All models on B05 ↗TECHNICAL
100/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · PLAYABILITY
68/100Band: Acceptable
FINAL
80/100Blend 87.2, capped at 80
Strong submission with 43 output files, 3201 LOC, and 37/37 automated Playwright QA checks passing. Official Oliver assets (18 PNGs + manifest) are integrated. All 6 core domain features detected. Self-contained with no CDN or eval. 58 build screenshots provide visual evidence. Minor gaps remain in confirming interactive game feel and precise thermal edge-case behavior without live playtesting.
- Time
- 55 min
- Tokens
- 46,797,708
- Client / effort
- MiniMax MCode
- Finished
- 23 October 2026
- Brief 1.0 · prompt SHA-256
- 9cf7e115eccf42ef9ad96ed7d7dc4ae064af4e783df7156c8601cf5189b14728
B06 / Wildcard One
All models on B06 ↗
CREATIVE SCORE
90/100Scored by the reviewer
- Time
- 1 h 10 min
- Client / effort
- MiniMax MCode
- Finished
- 23 October 2026
- Brief 1.0 · prompt SHA-256
- 827031e6b27e4e8d209db933a2e3a8166476216433fe01c2a1fd4083be4ebfbd
B07 / Wildcard Two
All models on B07 ↗CREATIVE SCORE
60/100Scored by the reviewer
- Client / effort
- M Code / High
- Brief 1.0 · prompt SHA-256
- e2555fa49a4563e27a174b1bef88a31085c0295186d7a4093e1f9832bf5c97c5