SVG: exploded French press, by claude-haiku-5.5
claude-haiku-5.5 · round 2 of 2 · seed 2$0.00SVGPNGpress.svg
press.svg · 91 lines
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 1000 1000" width="1000" height="1000">
  <defs>
    <pattern id="mesh" width="8" height="8" patternUnits="userSpaceOnUse">
      <path d="M0 0 H8 M0 0 V8" stroke="#777" stroke-width="1" fill="none"/>
    </pattern>
    <linearGradient id="glass" x1="0" x2="1">
      <stop offset="0" stop-color="#dfeef5"/>
      <stop offset="0.5" stop-color="#f7fbfd"/>
      <stop offset="1" stop-color="#d0e6f0"/>
    </linearGradient>
    <linearGradient id="steel" x1="0" x2="1">
      <stop offset="0" stop-color="#9aa3ab"/>
      <stop offset="0.5" stop-color="#e4e9ed"/>
      <stop offset="1" stop-color="#8b949c"/>
    </linearGradient>
  </defs>
  <rect width="1000" height="1000" fill="#ffffff"/>
  <text x="500" y="40" text-anchor="middle" font-family="Helvetica, Arial, sans-serif" font-size="26" font-weight="bold" fill="#222">French Press (three-piece plunger filter) – Exploded Assembly</text>

  <!-- alignment axis -->
  <line x1="400" y1="95" x2="400" y2="935" stroke="#999" stroke-width="1.5" stroke-dasharray="6 5"/>

  <!-- 1. Knob -->
  <g id="knob">
    <rect x="335" y="82" width="130" height="34" rx="16" fill="#2b2b2b"/>
    <rect x="365" y="116" width="70" height="20" rx="6" fill="#3a3a3a"/>
  </g>

  <!-- 2. Plunger rod -->
  <g id="rod">
    <rect x="393" y="136" width="14" height="172" fill="url(#steel)" stroke="#555" stroke-width="1"/>
  </g>

  <!-- 3. Lid (cap with spout hole) -->
  <g id="lid">
    <rect x="300" y="340" width="200" height="22" rx="10" fill="#2b2b2b"/>
    <rect x="382" y="345" width="36" height="12" fill="#ffffff"/>
    <rect x="390" y="332" width="20" height="14" fill="#444"/>
  </g>

  <!-- 4. Carafe (glass) with handle -->
  <g id="carafe">
    <path d="M474 440 C 560 430, 575 610, 474 606" fill="none" stroke="#333" stroke-width="16" stroke-linecap="round"/>
    <path d="M322 400 V 630 Q 322 652 344 652 H 456 Q 478 652 478 630 V 400 Z" fill="url(#glass)" stroke="#557" stroke-width="2" opacity="0.95"/>
    <path d="M478 400 Q 478 390 400 390 Q 322 390 322 400 Q 322 412 400 412 Q 478 412 478 400" fill="none" stroke="#557" stroke-width="1.5"/>
    <path d="M330 440 V 620" stroke="#fff" stroke-width="5" opacity="0.7"/>
  </g>

  <!-- 5. Filter holder (frame) -->
  <g id="holder">
    <rect x="312" y="690" width="176" height="18" rx="6" fill="#4a4a4a"/>
    <rect x="352" y="708" width="96" height="28" fill="#4a4a4a"/>
    <path d="M330 708 H470 M330 720 H470" stroke="#8a8a8a" stroke-width="2"/>
    <rect x="392" y="700" width="16" height="10" fill="#2b2b2b"/>
  </g>

  <!-- 6. Mesh screen -->
  <g id="screen">
    <path d="M318 770 V 880 Q 318 900 340 900 H 460 Q 482 900 482 880 V 770 Z" fill="url(#mesh)" stroke="#666" stroke-width="2"/>
    <ellipse cx="400" cy="770" rx="82" ry="10" fill="#e9edf0" stroke="#666" stroke-width="2"/>
  </g>

  <!-- alignment dashes at part boundaries -->
  <g stroke="#bbb" stroke-width="1" stroke-dasharray="3 4">
    <line x1="240" y1="122" x2="560" y2="122"/>
    <line x1="240" y1="330" x2="560" y2="330"/>
    <line x1="240" y1="390" x2="560" y2="390"/>
    <line x1="240" y1="676" x2="560" y2="676"/>
    <line x1="240" y1="760" x2="560" y2="760"/>
  </g>

  <!-- labels with leader lines -->
  <g font-family="Helvetica, Arial, sans-serif" font-size="18" fill="#222">
    <line x1="470" y1="99" x2="650" y2="99" stroke="#555"/>
    <text x="660" y="105">1  Knob (plunger handle)</text>
    <line x1="410" y1="222" x2="650" y2="222" stroke="#555"/>
    <text x="660" y="228">2  Plunger rod (stainless)</text>
    <line x1="500" y1="351" x2="650" y2="351" stroke="#555"/>
    <text x="660" y="357">3  Lid (with spout hole)</text>
    <line x1="520" y1="520" x2="650" y2="520" stroke="#555"/>
    <text x="660" y="526">4  Carafe (heat-resistant glass)</text>
    <line x1="490" y1="700" x2="650" y2="700" stroke="#555"/>
    <text x="660" y="706">5  Filter holder (frame)</text>
    <line x1="484" y1="830" x2="650" y2="830" stroke="#555"/>
    <text x="660" y="836">6  Mesh filter screen</text>
  </g>

  <!-- assembly note -->
  <text x="500" y="960" text-anchor="middle" font-family="Helvetica, Arial, sans-serif" font-size="16" fill="#555">Assembly: screen seats on holder, rod screws into holder, plunger passes through lid into carafe.</text>
</svg>
round 1 of 2
Exploded French press, by Claude Haiku 5.5, the render after round 1
12What the model saw after each round. The picture above is the round that was scored, round 2.
gallery/Exploded French press/Claude Haiku 5.5, run 2

Exploded French press: Claude Haiku 5.5

modelanthropic/claude-haiku-5.5tasksvg-french-press@v2, checks c1, harness h1date2026-10-08rounds2 of 8 allowed, 12 minutes on the clockcost$0.0040thinkingmediumtokens28,582 in, 5,340 out, of which 2,161 reasoningtime33 sfailedAll seven parts: incomplete, component missing, major, its note said nothing was wrong · Stack order: false, relation transposed, critical, near miss: 1 pair out of order, its note said nothing was wrong
All seven partsdecided by looking · Pass if all seven are drawn as separate shapes and each has a label naming it. A part that is only named in a label or a comment, with nothing drawn, fails it.fail
Stack orderdecided by a number, from the filefail
Filter fits the glassdecided by a number, from the filepass
One assembly axisdecided by looking · Pass if one vertical line runs through the middle of every part, and every leader line leaves from the same side. One part pushed off to the side fails it.pass
Labels matchdecided by looking · Pass if following each leader line from its label ends on the part that label names. One label pointing at the wrong part, or at nothing, fails it.pass

There is no post about this one, since only some results need one, so the picture, the source and the scoring are here instead.

its other runs
the prompt it was given
Draw an exploded assembly diagram of a French press, as an SVG. Use the standard three-piece plunger filter rather than a two-piece travel press.
Frozen at v2. This model got this and nothing else.
ground truth

The standard three-piece plunger, from the top down: spiral (spring) plate, mesh screen, cross-plate. The cross-plate is the bottom piece, because its threaded stem is what the rod screws into. The spiral plate above the mesh is what presses the mesh edge out against the glass. Checked against the WebstaurantStore anatomy diagram, which is drawn exploded and labelled, and against each part's y coordinate in the SVG rather than its comment. Two-piece travel presses exist and stack differently, which is why v2 of the prompt names the design. How it is scored, and this run's files on GitHub.