partforge

Drawer divider

hard tier · create

The request

Make a parametric drawer divider: a flat plate `length` mm along X, 40 mm tall along Z and 2 mm thick along Y, centred on the origin, with Z up. Cut `slots` equal slots down from the top edge, each 2.2 mm wide and 20 mm deep, evenly spaced along the length: the distance from each end of the plate to the nearest slot's centre and the distance between neighbouring slot centres are all equal, so the slots divide the length into `slots` + 1 equal steps. Expose two controls named exactly `length` (100 to 400 mm, step 10, default 200) and `slots` (2 to 8, step 1, default 4). Name the sub-part exactly `divider`.

What the AI judge checks

  • A long thin flat plate
  • Several narrow slots cut down from the top edge, all the same depth
  • The slots are evenly spaced with equal solid ends
  • Nothing else was added to the plate

Measured checks

  • one sub-part named divider (gate)
  • length is exposed as 100 to 400 (gate)
  • slots is exposed as 2 to 8 (gate)
  • default plate is 200 x 2 x 40
  • every slot is open and evenly spaced, the ends are solid
  • each slot is 2.2 mm wide, with plate either side of it
  • each slot is 20 mm deep, with plate below it

Ranked by hand

One reviewer put one attempt from each model in order, best first, ties left tied.

  1. #1 tie
    Gemini 3.8 Flash
    Gemini 3.8 Flash's ranked attempt
  2. #1 tie
    DeepSeek V4.1 Flash
    DeepSeek V4.1 Flash's ranked attempt
  3. #1 tie
    GPT-6 Luna
    GPT-6 Luna's ranked attempt
  4. #1 tie
    Opus 5.5
    Opus 5.5's ranked attempt
  5. #1 tie
    GPT-6 Sol
    GPT-6 Sol's ranked attempt
  6. #6
    Kimi K3 Fast
    Kimi K3 Fast's ranked attempt

    not acceptable

  7. #7
    Kimi K3 (Moonshot)

    not acceptable

Parts marked not acceptable fall below the reviewer's acceptable line whatever their order. Ordering by hand decides Human rank and Human rate on the leaderboard; it never affects Overall or Strict.

Every model's attempts

This is an older run. See the current results.

Run hard-baseline-2026-09-232026-09-23commit 6edf19e0spend $212.67 + $43.09 AI judging, 15 unpricedstopped early (interrupted)Other runs: 2026-10-16T18-49-10-hard-compact-2026-10-16, 2026-09-10T19-44-23-baseline-2026-09, 2026-09-04T20-02-10-stage2-hard-cases, 2026-09-04T19-48-31-screen-14-fixed