PhysDesignBench
GitHub
Design task · v2

MZI-Ring-v2

MZI-coupled microring add-drop filter electrothermal design v2

Design a fifth-order four-port Vernier microring add-drop filter with two physical bus MZIs, seven heaters, two fixed VOAs, and trusted local electrothermal and optical scoring.

Runs
13
Scored finals
7 of 13
Best final
S = 0.02 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) 0.016; Opus 5.5 (Claude Code) 0.010; Fable 5.1 (Claude Code) 0.010; GPT-6.1 Sol (mini-SWE-agent) 0.010; Sonnet 5.5 (mini-SWE-agent) 0.009; GPT-6.1 Sol (Codex) 0.009; Muse Wonder (high) (Muse Code) 0.007; Gemini 4 Argon (mini-SWE-agent) Invalid; GLM 5.3 Flash (high) (Claude Code) Invalid; GLM 5.3 Flash (max) (mini-SWE-agent) Invalid; GPT-5.6 Sol (Codex) Invalid; Kimi K3.1 Pro (Claude Code) Invalid; Sonnet 5.5 (high) (Claude Code) Invalid.0.000.250.500.751.000153045607590105120135Elapsed time (min)Score SSonnet 5.5 · mini-SWE-agent: final S = 0.009 at 107 minGPT-6.1 Sol · mini-SWE-agent: final S = 0.010 at 122 minGemini 4 Argon · mini-SWE-agent: final Invalid at 121 minGLM 5.3 Flash (high) · Claude Code: final Invalid at 121 minGLM 5.3 Flash (max) · mini-SWE-agent: final Invalid at 121 minGPT-5.6 Sol · Codex: final Invalid at 121 minKimi K3.1 Pro · Claude Code: final Invalid at 100 minSonnet 5.5 (high) · Claude Code: final Invalid at 58 minMuse Wonder (high) · Muse Code: final S = 0.007 at 105 minGPT-6.1 Sol · Codex: final S = 0.009 at 121 minFable 5.1 · Claude Code: final S = 0.010 at 100 minOpus 5.5 · Claude Code: final S = 0.010 at 49 minGPT-6 Astra · Codex: final S = 0.016 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
  • VOA marker
  • Layer 35/0
  • Path marker
  • Layer 37/0

  • MZI-Ring-v2 · Codex · GPT-6 Astra

    S = 0.016 · scale bar 100 µm

    Fit
    GPT-6 AstraCodexS = 0.02

  • MZI-Ring-v2 · Claude Code · Claude Opus 5.5

    S = 0.010 · scale bar 100 µm

    Fit
    Opus 5.5Claude CodeS = 0.01

  • MZI-Ring-v2 · Claude Code · Claude Fable 5.1

    S = 0.010 · scale bar 100 µm

    Fit
    Fable 5.1Claude CodeS = 0.01

  • MZI-Ring-v2 · mini-SWE-agent · GPT-6.1-sol

    S = 0.010 · scale bar 100 µm

    Fit
    GPT-6.1 Solmini-SWE-agentS = 0.01

  • MZI-Ring-v2 · mini-SWE-agent · Claude Sonnet 5.5

    S = 0.009 · scale bar 100 µm

    Fit
    Sonnet 5.5mini-SWE-agentS = 0.01

  • MZI-Ring-v2 · Codex · GPT-6.1-sol

    S = 0.009 · scale bar 100 µm

    Fit
    GPT-6.1 SolCodexS = 0.01

  • MZI-Ring-v2 · Muse Code · Muse Wonder

    S = 0.007 · scale bar 100 µm

    Fit
    Muse Wonder (high)Muse CodeS = 0.01

  • MZI-Ring-v2 · mini-SWE-agent · Gemini-4-Argon

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    Gemini 4 Argonmini-SWE-agentInvalid · S = 0
  • No final GDS: the evaluator rejected the final source before producing a layout
    GLM 5.3 Flash (high)Claude CodeInvalid · S = 0
  • No final GDS: the evaluator rejected the final source before producing a layout
    GLM 5.3 Flash (max)mini-SWE-agentInvalid · S = 0

  • MZI-Ring-v2 · Codex · GPT-5.6-sol

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    GPT-5.6 SolCodexInvalid · S = 0

  • MZI-Ring-v2 · Claude Code · Kimi K3.1 Pro

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    Kimi K3.1 ProClaude CodeInvalid · S = 0

  • MZI-Ring-v2 · Claude Code · Claude Sonnet 5.5

    Invalid final · S = 0 · scale bar 100 µm

    Fit
    Sonnet 5.5 (high)Claude CodeInvalid · S = 0