← mecheval / task / a1-stepped-shaft-01

Stepped shaft A · A1 · a1-stepped-shaft-01

primitives · boolean · axisymmetric

Expected

Prompt

Make a stepped shaft along the Z axis: a base cylinder of diameter 40mm and height 10mm sitting on the XY plane (z = 0 to z = 10), with a smaller coaxial cylinder of diameter 20mm and height 15mm stacked directly on top (z = 10 to z = 25). The result is a single solid.

Checks

0
valid_solid
{
  "type": "valid_solid"
}
1
bbox
{
  "type": "bbox",
  "min": [
    -20,
    -20,
    0
  ],
  "max": [
    20,
    20,
    25
  ],
  "tolerance_mm": 0.1
}
2
mass_props
{
  "type": "mass_props",
  "volume_mm3": 17278.76,
  "tolerance_pct": 1
}
3
step_roundtrip
{
  "type": "step_roundtrip",
  "tolerance_pct": 1
}

Anti-cheese

{
  "max_solid_count": 1
}

Limits

{
  "max_tokens": 30000,
  "max_wallclock_sec": 180,
  "max_tool_calls": 30
}

Recent attempts

Runs (41)

modelrun statusscorefirst failtokenswall
claude-mcp-claude-opus-4-7 20260611T172300Z-38d8 PASS 1.00 365.7k 72.3s
claude-mcp-claude-opus-4-7 20260611T172252Z-4bf8 PASS 1.00 203.4k 44.9s
claude-mcp-claude-opus-4-7 20260611T172153Z-e84b PASS 1.00 254.0k 59.3s
claude-mcp-claude-opus-4-7 20260611T172147Z-b479 PASS 1.00 416.2k 72.3s
claude-mcp-claude-opus-4-7 20260611T172144Z-310e PASS 1.00 405.2k 77.0s
openai-direct-gpt-5-mini 20260428T212831Z-72f3 PASS 1.00 1.5k 12.5s
openai-direct-gpt-5-mini 20260428T212818Z-4f83 PASS 1.00 1.5k 13.4s
openai-direct-gpt-5-mini 20260428T212756Z-e826 PASS 1.00 1.3k 21.4s
openai-direct-gpt-5-mini 20260428T212744Z-6ce8 PASS 1.00 1.3k 12.3s
openai-direct-gpt-5-mini 20260428T212732Z-d512 PASS 1.00 1.3k 12.1s
openai-direct-gpt-5 20260428T212708Z-c37e PASS 1.00 1.4k 8.8s
openai-direct-gpt-5 20260428T212657Z-ddcf PASS 1.00 1.4k 11.5s
openai-direct-gpt-5 20260428T212646Z-5538 PASS 1.00 1.4k 10.9s
openai-direct-gpt-5 20260428T212636Z-5690 PASS 1.00 1.4k 10.4s
openai-direct-gpt-5 20260428T212626Z-2d9a PASS 1.00 1.3k 9.3s
openai-direct-gpt-4o-mini 20260428T212535Z-e73e PASS 1.00 836 3.5s
openai-direct-gpt-4o-mini 20260428T212531Z-59c3 fail 0.00 valid_solid · solid invalid 861 4.1s
openai-direct-gpt-4o-mini 20260428T212526Z-53ae fail 0.00 valid_solid · solid invalid 861 4.9s
openai-direct-gpt-4o-mini 20260428T212522Z-1bc6 fail 0.00 valid_solid · solid invalid 866 4.1s
openai-direct-gpt-4o-mini 20260428T212518Z-05e6 fail 0.00 valid_solid · solid invalid 865 4.0s
claude-direct-claude-sonnet-4-6 20260428T211358Z-0d14 PASS 1.00 920 3.3s
claude-direct-claude-sonnet-4-6 20260428T211354Z-a26a PASS 1.00 927 3.9s
claude-direct-claude-sonnet-4-6 20260428T211351Z-b805 PASS 1.00 926 3.4s
claude-direct-claude-sonnet-4-6 20260428T211347Z-263f PASS 1.00 927 4.0s
claude-direct-claude-sonnet-4-6 20260428T211343Z-7be8 PASS 1.00 925 3.9s
claude-direct-claude-haiku-4-5-20251001 20260428T211340Z-356d fail 0.00 valid_solid · solid invalid 1.0k 2.5s
claude-direct-claude-haiku-4-5-20251001 20260428T211337Z-0510 fail 0.00 valid_solid · solid invalid 1.1k 2.4s
claude-direct-claude-haiku-4-5-20251001 20260428T211335Z-3895 fail 0.00 valid_solid · solid invalid 1.1k 2.5s
claude-direct-claude-haiku-4-5-20251001 20260428T211332Z-40dc fail 0.50 bbox · X off by -20.00mm 1.1k 2.6s
claude-direct-claude-haiku-4-5-20251001 20260428T211330Z-227a PASS 1.00 1.0k 2.1s
claude-direct-claude-sonnet-4-6 20260428T211208Z-baab PASS 1.00 937 3.8s
claude-direct-claude-sonnet-4-6 20260428T211205Z-7b19 PASS 1.00 926 3.7s
claude-direct-claude-sonnet-4-6 20260428T211201Z-c8dc PASS 1.00 927 3.7s
claude-direct-claude-sonnet-4-6 20260428T211157Z-078c PASS 1.00 927 3.8s
claude-direct-claude-sonnet-4-6 20260428T211153Z-cd47 PASS 1.00 926 3.9s
claude-direct-claude-opus-4-7 20260428T153537Z-33d7 PASS 1.00 1.2k 3.4s
claude-direct-claude-opus-4-7 20260428T153525Z-8710 PASS 1.00 1.1k 5.0s
claude-direct-claude-opus-4-7 20260428T153513Z-ab32 PASS 1.00 1.2k 3.4s
claude-mcp-claude-opus-4-7 20260428T145724Z-9051 PASS 1.00 402.2k 69.4s
claude-direct-claude-opus-4-7 20260428T142500Z-3bc3 PASS 1.00 1.2k 3.6s
claude-direct-claude-opus-4-7 20260428T141639Z-92f3 PASS 1.00 1.2k 3.9s

generated 2026-06-17T03:16:07.199Z · static site, regenerate with npm run build -w @mecheval/leaderboard