MODEL GAUNTLET / COMPETITOR PROFILE
Beam 1.0 501B A23B.
Rank 4 of 4 · 289.8/500 across B01–B05 · Wildcards 50/200.
B01 / Mobile OS
All models on B01 ↗TECHNICAL
88.5/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · FUNCTIONALITY
76/100Band: Acceptable
FINAL
80/100Blend 83.5, capped at 80
Vesper OS (B01-v1.0) is a self-contained single-page mobile OS simulation with all 8 core domain features detected in 1,351 LOC across 4 files, launching cleanly with zero external dependencies. QA documentation confirms lock-screen unlock, messaging, notes, and calculator functionality, but several features (phone dialer, contacts CRUD, settings persistence, gallery) lack interactive test evidence.
- Time
- 9 min
- Tokens
- 4,899,867
- Client / effort
- OpenCode / max
- Finished
- 21 October 2026
- Brief 1.0 · prompt SHA-256
- a00e95f7a9243ba2b2a9046d9ff7225c565196138c18322f4a0c659241bf9755
B02 / Blender Mining Drone
All models on B02 ↗
TECHNICAL
70/100Judged by qwen3.8-flash-next · ceiling 70
HUMAN · VISUAL QUALITY
20/100Band: Weak
FINAL
50/100Not capped
Solid procedural Blender mining drone with 128 meshes, 3642 triangles, 8 PBR materials, and all 7 required subsystems detected. GLB exports cleanly with proper pivots and unit transforms. Main weaknesses: GLB node naming is primitive (Cube.001), no bevel/sharp-edge control evident, and the model self-admits it cannot verify visual fidelity to the reference image.
- Time
- 16 min
- Tokens
- 2,530,484
- Client / effort
- OpenCode / max
- Finished
- 21 October 2026
- Brief 1.0 · prompt SHA-256
- 4d9e8edee96c488962673f7c21868a8cba3ea55012bbef6c9e733f505dd413b3
B03 / Blender to Godot
All models on B03 ↗TECHNICAL
81/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · PLAYABILITY
52/100Band: Weak
FINAL
60/100Blend 69.4, capped at 60
Solid B03 submission with exact B02 GLB provenance (SHA-256 match), clean Godot 4.7.2 project launching without errors, and complete gameplay code covering movement, mining, win state, and UI. The main gap is absence of live playtest evidence (no Godot runtime screenshots or video); all functional claims rest on headless script tests and static code inspection. Asset fidelity, project structure, and documentation are strong.
- Time
- 28 min
- Tokens
- 4,705,780
- Client / effort
- OpenCode / max
- Finished
- 21 October 2026
- Brief 1.0 · prompt SHA-256
- f00af20d56033bae25fc2d36a139ec5f56e08b50d99be07946ad19c06a5f5527
B04 / Prime London Estate Agent
All models on B04 ↗TECHNICAL
68.5/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · DESIGN & NAVIGATION
46.7/100Band: Acceptable
FINAL
59.8/100Not capped
Hartwell Prime Property is a well-structured React/TypeScript/Vite/Tailwind estate agency site with all 8 core feature signals detected, 11/11 tests passing, and all routes returning 200. The codebase demonstrates URL-based search state, localStorage shortlist persistence, form validation with zod, SVG floor plans, and a full-screen gallery.
- Time
- 7 min
- Tokens
- 6,032,776
- Client / effort
- OpenCode / max
- Finished
- 23 October 2026
- Brief 2.0 · prompt SHA-256
- b2229dc93c7a8d4ec7dda8712ff98e2f0c8cbc618181cd8cc2239ac458d1bc50
B05 / Oliver GPU Game
All models on B05 ↗TECHNICAL
68/100Judged by qwen3.8-flash-next · ceiling 100
HUMAN · PLAYABILITY
0/100Band: Unusable
FINAL
40/100Blend 40.8, capped at 40
A compact 484-LOC 2D canvas game with verified thermal simulation (94.7C→70.7C cooling delta), CDP-tested state transitions, and official Oliver sprite integration. Strong code-level evidence for core mechanics but zero screenshots/video prevents visual confirmation of HUD legibility, game feel, animation states, and visual polish. The submission is functionally complete per QA documentation but lacks the visual evidence needed to confirm arcade-quality presentation.
- Time
- 6 min
- Tokens
- 5,390,853
- Client / effort
- OpenCode / max
- Finished
- 21 October 2026
- Brief 1.0 · prompt SHA-256
- 9cf7e115eccf42ef9ad96ed7d7dc4ae064af4e783df7156c8601cf5189b14728
B06 / Wildcard One
All models on B06 ↗
CREATIVE SCORE
40/100Scored by the reviewer
- Time
- 10 min
- Tokens
- 2,892,604
- Client / effort
- OpenCode / max
- Finished
- 21 October 2026
- Brief 1.0 · prompt SHA-256
- 827031e6b27e4e8d209db933a2e3a8166476216433fe01c2a1fd4083be4ebfbd
B07 / Wildcard Two
All models on B07 ↗CREATIVE SCORE
10/100Scored by the reviewer
- Time
- 12 min
- Tokens
- 4,054,908
- Client / effort
- OpenCode / max
- Finished
- 21 October 2026
- Brief 1.0 · prompt SHA-256
- e2555fa49a4563e27a174b1bef88a31085c0295186d7a4093e1f9832bf5c97c5