Mesh holder holes

Kimi K3.1 Pro$0.181m 29s1m 25s waiting on the model2 builds

Reference
iso
front
back
top
bottom
left
right

95of 100

built: 100
geometry: 91
AI judge: 7 of 7

What the AI judge checked

Pass:The overall form matches the reference: the same kind of object, the same silhouette from the matching angle, allowing for the requested change: the holder scaled up ten times to real size, otherwise unchanged, with two through-holes added in the floor of the channel
- Opus 5.5: pass: The result shows the same channel-shaped holder with bowed side walls and an open-ended slot, matching the reference silhouettes from every angle.
- GPT-6 Sol: pass: The result has the same open-channel holder silhouette as the reference, with two holes added in the channel floor.
Pass:Proportions match: the relative sizes of the major dimensions agree with the reference within what the eye can judge, allowing for the requested change: the holder scaled up ten times to real size, otherwise unchanged, with two through-holes added in the floor of the channel
- Opus 5.5: pass: The length, height and width ratios in the front, end and top views match the reference renders.
- GPT-6 Sol: pass: The length, height, wall thickness, and channel width appear proportionally consistent with the reference.
Pass:Every major feature visible in the reference (holes, slots, bosses, tabs, lips) is present in the result, allowing for the requested change: the holder scaled up ten times to real size, otherwise unchanged, with two through-holes added in the floor of the channel
- Opus 5.5: pass: The channel, tapered slot mouth, central V-notches in the walls and rounded edges all appear as in the reference.
- GPT-6 Sol: pass: The open channel, bowed side walls, rounded edges, and central wall contours visible in the reference are present.
Pass:Those features sit where the reference has them, allowing for the requested change: the holder scaled up ten times to real size, otherwise unchanged, with two through-holes added in the floor of the channel
- Opus 5.5: pass: The notches sit at mid-length and the channel is centred, exactly as in the reference.
- GPT-6 Sol: pass: The channel and its wall contours occupy the same relative positions as in the reference.
Pass:Nothing is present that the reference does not show — no decorative or unexplained geometry, allowing for the requested change: the holder scaled up ten times to real size, otherwise unchanged, with two through-holes added in the floor of the channel
- Opus 5.5: pass: Apart from the two requested holes, no extra geometry appears compared with the reference.
- GPT-6 Sol: pass: No unexplained geometry is visible beyond the two requested holes.
Pass:The holder is the imported shape unchanged — the slightly bowed side walls and the open channel are as in the reference — apart from the two holes
- Opus 5.5: pass: The bowed side walls and the open channel are identical to the reference in all views.
- GPT-6 Sol: pass: Apart from the holes, the renders show the reference’s slightly bowed walls and open channel without a visible shape change.
Pass:Two round holes go through the floor of the channel on its centreline, one near each end, and nothing else was cut
- Opus 5.5: pass: The top and bottom views show two round through-holes on the channel centreline, roughly a quarter of the length in from each end, and nothing else is cut.
- GPT-6 Sol: pass: Two round holes are visible on the channel floor centreline near opposite ends, and the underside view shows both openings.

Measured

✓ watertight: true == true
✓ one sub-part named holder: ok
✓ the imported holder is still declared as an import: ok
✓ the mesh was scaled to real size: ok
✓ the holder's own volume survives: ok
✓ nothing solid was added inside the holder's own footprint: ok
✓ the +Y hole clears 4.5 mm through the floor: ok
✓ the -Y hole clears 4.5 mm through the floor: ok
✓ the floor is still there around the +Y hole: ok
✓ the floor is still there around the -Y hole: ok
✗ measured the mesh with the oracle: predicate returned false

Loading the transcript…

This is an older run. See the current results.

Run kimi31-sonnet-hard2026-11-05commit 58201d93spend $50.99 + $14.79 AI judgingOther runs: 2026-11-05T05-50-12-kimi31-vs-sonnet, 2026-10-16T18-49-10-hard-compact-2026-10-16, 2026-09-23T06-56-16-hard-baseline-2026-09-23, 2026-09-10T19-44-23-baseline-2026-09, 2026-09-04T20-02-10-stage2-hard-cases, 2026-09-04T19-48-31-screen-14-fixed