SVG: exploded AeroPress, by claude-haiku-5.5
claude-haiku-5.5 · round 3 of 3 · seed 1$0.01SVGPNGaeropress.svg
aeropress.svg · 95 lines
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 800 1000" width="800" height="1000" font-family="Helvetica, Arial, sans-serif">
  <defs>
    <linearGradient id="clear" x1="0" x2="1" y1="0" y2="0">
      <stop offset="0" stop-color="#dfeef5"/>
      <stop offset="0.35" stop-color="#ffffff"/>
      <stop offset="1" stop-color="#cfe3ee"/>
    </linearGradient>
    <linearGradient id="plastic" x1="0" x2="1" y1="0" y2="0">
      <stop offset="0" stop-color="#b8bcc2"/>
      <stop offset="0.4" stop-color="#eef0f2"/>
      <stop offset="1" stop-color="#9ea4ab"/>
    </linearGradient>
    <linearGradient id="rubber" x1="0" x2="1" y1="0" y2="0">
      <stop offset="0" stop-color="#2b2b2b"/>
      <stop offset="0.4" stop-color="#555"/>
      <stop offset="1" stop-color="#1e1e1e"/>
    </linearGradient>
    <linearGradient id="paper" x1="0" x2="1" y1="0" y2="0">
      <stop offset="0" stop-color="#c9b48a"/>
      <stop offset="0.5" stop-color="#f3e6c4"/>
      <stop offset="1" stop-color="#c2aa7c"/>
    </linearGradient>
    <linearGradient id="cap" x1="0" x2="1" y1="0" y2="0">
      <stop offset="0" stop-color="#4f6f8c"/>
      <stop offset="0.45" stop-color="#8fb1cc"/>
      <stop offset="1" stop-color="#3f5d78"/>
    </linearGradient>
    <pattern id="mesh" width="10" height="10" patternUnits="userSpaceOnUse">
      <circle cx="5" cy="5" r="2" fill="#5d6770"/>
    </pattern>
  </defs>

  <rect width="800" height="1000" fill="#ffffff"/>
  <text x="400" y="34" text-anchor="middle" font-size="24" font-weight="bold" fill="#222">AeroPress — Exploded Assembly</text>
  <text x="400" y="56" text-anchor="middle" font-size="13" fill="#666">Original model with round paper filter</text>

  <!-- alignment axis -->
  <line x1="400" y1="75" x2="400" y2="935" stroke="#bbb" stroke-width="1" stroke-dasharray="6,5"/>

  <!-- 1. PLUNGER -->
  <rect x="322" y="92" width="156" height="34" rx="16" fill="url(#plastic)" stroke="#555" stroke-width="1.5"/>
  <rect x="386" y="126" width="28" height="206" fill="url(#plastic)" stroke="#555" stroke-width="1.5"/>
  <!-- 2. RUBBER SEAL -->
  <rect x="346" y="332" width="108" height="30" rx="6" fill="url(#rubber)" stroke="#111" stroke-width="1.5"/>
  <line x1="346" y1="342" x2="454" y2="342" stroke="#777" stroke-width="1"/>
  <line x1="346" y1="352" x2="454" y2="352" stroke="#777" stroke-width="1"/>

  <!-- 3. CHAMBER -->
  <rect x="344" y="446" width="112" height="250" fill="url(#clear)" stroke="#4a6b7d" stroke-width="2"/>
  <rect x="330" y="430" width="140" height="18" rx="6" fill="url(#clear)" stroke="#4a6b7d" stroke-width="2"/>
  <rect x="334" y="694" width="132" height="14" rx="4" fill="#dfeef5" stroke="#4a6b7d" stroke-width="2"/>
  <g stroke="#4a6b7d" stroke-width="1">
    <line x1="344" y1="500" x2="362" y2="500"/>
    <line x1="344" y1="550" x2="356" y2="550"/>
    <line x1="344" y1="600" x2="362" y2="600"/>
    <line x1="344" y1="650" x2="356" y2="650"/>
  </g>

  <!-- 4. PAPER FILTER -->
  <ellipse cx="400" cy="770" rx="60" ry="9" fill="url(#paper)" stroke="#8a7650" stroke-width="1.5"/>
  <ellipse cx="400" cy="767" rx="60" ry="9" fill="#f7eed8" stroke="#8a7650" stroke-width="1"/>

  <!-- 5. FILTER CAP -->
  <ellipse cx="400" cy="906" rx="62" ry="10" fill="url(#mesh)" stroke="#44515c" stroke-width="1.5"/>
  <rect x="340" y="812" width="120" height="88" fill="url(#cap)" stroke="#2c4157" stroke-width="1.5"/>
  <g stroke="#2c4157" stroke-width="1.5" opacity="0.6">
    <line x1="356" y1="818" x2="356" y2="894"/>
    <line x1="372" y1="818" x2="372" y2="894"/>
    <line x1="388" y1="818" x2="388" y2="894"/>
    <line x1="404" y1="818" x2="404" y2="894"/>
    <line x1="420" y1="818" x2="420" y2="894"/>
    <line x1="436" y1="818" x2="436" y2="894"/>
  </g>
  <rect x="332" y="802" width="136" height="14" rx="5" fill="url(#cap)" stroke="#2c4157" stroke-width="1.5"/>

  <!-- LABEL LEADERS (right side only) -->
  <g stroke="#333" stroke-width="1">
    <line x1="480" y1="109" x2="590" y2="109"/>
    <line x1="456" y1="350" x2="590" y2="350"/>
    <line x1="456" y1="500" x2="590" y2="500"/>
    <line x1="462" y1="767" x2="590" y2="767"/>
    <line x1="462" y1="860" x2="590" y2="860"/>
  </g>
  <g font-size="15" fill="#222">
    <text x="596" y="105">1  Plunger</text>
    <text x="596" y="346">2  Rubber seal</text>
    <text x="596" y="496">3  Chamber</text>
    <text x="596" y="763">4  Paper filter (round)</text>
    <text x="596" y="856">5  Filter cap</text>
  </g>

  <text x="400" y="955" text-anchor="middle" font-size="13" fill="#666">Parts shown exploded along the assembly axis.</text>
  <text x="400" y="973" text-anchor="middle" font-size="13" fill="#666">Assembly: place paper filter in filter cap, screw cap onto chamber base, then insert plunger (seal on) from the top.</text>
</svg>
round 1 of 3
Exploded AeroPress, by Claude Haiku 5.5, the render after round 1
123What the model saw after each round. The picture above is the round that was scored, round 3.
gallery/Exploded AeroPress/Claude Haiku 5.5, run 1

Exploded AeroPress: Claude Haiku 5.5

modelanthropic/claude-haiku-5.5taskexploded-aeropress@v1, checks c1, harness h1date2026-10-08rounds3 of 8 allowed, 12 minutes on the clockcost$0.0085thinkingmediumtokens108,230 in, 11,568 out, of which 3,488 reasoningtime1 min 10 s
All five partsdecided by looking · Pass if every listed part is drawn as a separate shape and each has a label naming it. A part only named in a label or a comment, with nothing drawn, fails it. The label must say what the part is, in any words: a name that only says where it sits ("top plate") or what it is made of, or a mention in another part's caption, does not count.pass
Stack orderdecided by a number, from the filepass
Parts fitdecided by a number, from the filepass
One assembly axisdecided by looking · Pass if one line runs through the middle of every part on each listed axis, and every leader line leaves from the same side. One part pushed off its axis 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
Exploded AeroPress, by Claude Haiku 5.5, run 2run 2 · 2r · $0.0063Exploded AeroPress, by Claude Haiku 5.5, run 3run 3 · 2r · $0.0056
same prompt, other modelsall 15 runs on Exploded AeroPress →
the prompt it was given
Draw an exploded assembly diagram of an AeroPress coffee maker, as an SVG. Use the original model with a round paper filter rather than the Go or the XL.
Frozen at v1. This model got this and nothing else.
ground truth

Top to bottom: plunger, rubber seal, chamber, paper filter, filter cap. The seal sits on the plunger's lower end, and the paper filter sits in the cap below the chamber. Checked against AeroPress's replacement-parts pages, which sell the plunger with its seal, the chamber and the cap as the parts of the Original, and Wikipedia's description of the cap holding the filter at the chamber's bottom end. The stirrer and scoop are not part of the assembly. How it is scored, and this run's files on GitHub.