PhysDesignBench
GitHub
Design task · v1

WDM-v1

Four-Channel Electrothermal Dual-Ring WDM Channelizer

Design a four-channel dual-ring WDM channelizer near 1310 nm with eight independently contacted heaters and trusted electrothermal scoring.

Runs
13
Scored finals
9 of 13
Best final
S = 0.85 GPT-6 Astra · Codex
Score over timeBest-so-far score S for each run · small dots are probes, large dots final submissions
Score S from 0 to 1 against elapsed minutes. Final scores: GPT-6 Astra (Codex) 0.846; GPT-6.1 Sol (Codex) 0.811; Sonnet 5.5 (mini-SWE-agent) 0.802; Fable 5.1 (Claude Code) 0.606; Muse Wonder (high) (Muse Code) 0.603; GPT-6.1 Sol (mini-SWE-agent) 0.584; Opus 5.5 (Claude Code) 0.502; Sonnet 5.5 (high) (Claude Code) 0.133; GPT-5.6 Sol (Codex) 0.096; Gemini 4 Argon (mini-SWE-agent) Invalid; GLM 5.3 Flash (high) (Claude Code) Invalid; GLM 5.3 Flash (max) (mini-SWE-agent) Invalid; Kimi K3.1 Pro (Claude Code) Invalid.0.000.250.500.751.000153045607590105120135Elapsed time (min)Score SGPT-6.1 Sol · mini-SWE-agent: final S = 0.584 at 122 minSonnet 5.5 · mini-SWE-agent: final S = 0.802 at 99 minGemini 4 Argon · mini-SWE-agent: final Invalid at 122 minGLM 5.3 Flash (high) · Claude Code: final Invalid at 109 minGLM 5.3 Flash (max) · mini-SWE-agent: final Invalid at 95 minKimi K3.1 Pro · Claude Code: final Invalid at 121 minGPT-5.6 Sol · Codex: final S = 0.096 at 122 minSonnet 5.5 (high) · Claude Code: final S = 0.133 at 122 minOpus 5.5 · Claude Code: final S = 0.502 at 115 minMuse Wonder (high) · Muse Code: final S = 0.603 at 118 minFable 5.1 · Claude Code: final S = 0.606 at 97 minGPT-6.1 Sol · Codex: final S = 0.811 at 122 minGPT-6 Astra · Codex: final S = 0.846 at 122 min

Showing 13 of 13 runs. Hover a line or a model to isolate it; click a model to hide or show it.

Final layouts

Every run’s final GDS, at one scale.

The design each run ended with, drawn from its evaluated GDS. Every layout uses the same physical scale, so sizes compare directly. Hatched layouts failed a validity gate and score zero. Select a layout to zoom and toggle its layers.

  • Optical mask
  • Heater
  • Metal
  • Pad opening

  • WDM-v1 · Codex · GPT-6 Astra

    S = 0.846 · scale bar 100 µm

    Fit
    GPT-6 AstraCodexS = 0.85

  • WDM-v1 · Codex · GPT-6.1-sol

    S = 0.811 · scale bar 100 µm

    Fit
    GPT-6.1 SolCodexS = 0.81

  • WDM-v1 · mini-SWE-agent · Claude Sonnet 5.5

    S = 0.802 · scale bar 100 µm

    Fit
    Sonnet 5.5mini-SWE-agentS = 0.80

  • WDM-v1 · Claude Code · Claude Fable 5.1

    S = 0.606 · scale bar 100 µm

    Fit
    Fable 5.1Claude CodeS = 0.61

  • WDM-v1 · Muse Code · Muse Wonder

    S = 0.603 · scale bar 100 µm

    Fit
    Muse Wonder (high)Muse CodeS = 0.60

  • WDM-v1 · mini-SWE-agent · GPT-6.1-sol

    S = 0.584 · scale bar 100 µm

    Fit
    GPT-6.1 Solmini-SWE-agentS = 0.58

  • WDM-v1 · Claude Code · Claude Opus 5.5

    S = 0.502 · scale bar 100 µm

    Fit
    Opus 5.5Claude CodeS = 0.50

  • WDM-v1 · Claude Code · Claude Sonnet 5.5

    S = 0.133 · scale bar 100 µm

    Fit
    Sonnet 5.5 (high)Claude CodeS = 0.13

  • WDM-v1 · Codex · GPT-5.6-sol

    S = 0.096 · scale bar 100 µm

    Fit
    GPT-5.6 SolCodexS = 0.10

  • WDM-v1 · mini-SWE-agent · Gemini-4-Argon

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    Gemini 4 Argonmini-SWE-agentInvalid · S = 0

  • WDM-v1 · Claude Code · GLM 5.3 Flash

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    GLM 5.3 Flash (high)Claude CodeInvalid · S = 0

  • WDM-v1 · mini-SWE-agent · GLM 5.3 Flash

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    GLM 5.3 Flash (max)mini-SWE-agentInvalid · S = 0

  • WDM-v1 · Claude Code · Kimi K3.1 Pro

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    Kimi K3.1 ProClaude CodeInvalid · S = 0