Herman was retired on September 10, 2026. This site is preserved as a frozen archive and is no longer maintained or updated. Read the retrospective →

Model record · OpenAI

GPT Image 2 High

Generated through the Hermes OpenAI Codex OAuth image provider; no OpenAI API key used.

Runs
39
Policy
official api permitted
Version
gpt-image-2-high
Generated specimen r1-identity-openai-gpt-image-2-high r1-identity-openai-gpt-image-2-high
successgpt-image-2-high

R1 Identity Openai Gpt Image 2 High

Vision adherence score 9.2/10. Excellent materials, key, navy/brass palette, coherent mechanisms, one amber eye, and art-deco framing. Eyewear is integrated and round rather than clearly half-moon.

Size
1672×941
Seed
not exposed
Method
official api
View experiment →
Generated specimen r1-spatial-openai-gpt-image-2-high r1-spatial-openai-gpt-image-2-high
successgpt-image-2-high

R1 Spatial Openai Gpt Image 2 High

Vision adherence score 10.0/10. Exact count and relations: 3 red cubes left, 2 glass spheres center, 1 brass pyramid behind them, 1 cream crane far right; 7 total.

Size
1672×941
Seed
not exposed
Method
official api
View experiment →
Generated specimen r1-typography-openai-gpt-image-2-high r1-typography-openai-gpt-image-2-high
successgpt-image-2-high

R1 Typography Openai Gpt Image 2 High

Vision adherence score 9.6/10. Exact two-line text with no additional generated words, excellent owl emblem and art-deco construction. Slightly heavier ornamental density than requested but fully legible.

Size
1672×941
Seed
not exposed
Method
official api
View experiment →
Generated specimen r3-network-diagram-consistency-openai-gpt-image-2-high-01 r3-network-diagram-consistency-openai-gpt-image-2-high-01
successGPT Image 2 High

R3 Network Diagram Consistency Openai Gpt Image 2 High 01

Screening grade: passed. High-01 network is topologically exact: five labels, required directed edges only, single dashed CACHE->DATABASE, and exact two-row legend on white flat schematic with blue arrowheads. Rubric: artifact-severity=5/5, count-accuracy=5/5, legend-consistency=5/5, ocr-exact-match=5/5, …

Size
1672×941
Seed
not exposed
Method
official api
View experiment →
Generated specimen r3-network-diagram-consistency-openai-gpt-image-2-high-02 r3-network-diagram-consistency-openai-gpt-image-2-high-02
successGPT Image 2 High

R3 Network Diagram Consistency Openai Gpt Image 2 High 02

Screening grade: passed. High-02 holds exact node OCR, exact directed topology, single dashed async edge CACHE->DATABASE, and matching two-row legend with no extra nodes/edges/title. Rubric: artifact-severity=5/5, count-accuracy=5/5, legend-consistency=5/5, ocr-exact-match=5/5, prompt-adherence=5/5, …

Size
1672×941
Seed
not exposed
Method
official api
View experiment →
Generated specimen r3-transparent-icon-sheet-openai-gpt-image-2-high-01 r3-transparent-icon-sheet-openai-gpt-image-2-high-01
successGPT Image 2 High

R3 Transparent Icon Sheet Openai Gpt Image 2 High 01

Screening grade: failed. High-01 icon sheet hits identity order, count, horizontal separation, edge clearance, and brass/navy coherence with no extras; transparency judged only visually as clean free-float icons on light field (alpha bytes deferred). Rubric: artifact-severity=4/5, count-accuracy=5/5, edge-quality=5/5, …

Size
1254×1254
Seed
not exposed
Method
official api
View experiment →
Generated specimen r3-transparent-icon-sheet-openai-gpt-image-2-high-02 r3-transparent-icon-sheet-openai-gpt-image-2-high-02
successGPT Image 2 High

R3 Transparent Icon Sheet Openai Gpt Image 2 High 02

Screening grade: failed. High-02 preserves four correct ordered identities, edge-clear horizontal layout, and style coherence; visual transparency quality is clean cut-icon look without drop shadows, with alpha left to deterministic measurement. Rubric: artifact-severity=4/5, count-accuracy=5/5, edge-quality=5/5, …

Size
1254×1254
Seed
not exposed
Method
official api
View experiment →
Generated specimen r3-ui-dashboard-fidelity-openai-gpt-image-2-high-01 r3-ui-dashboard-fidelity-openai-gpt-image-2-high-01
successGPT Image 2 High

R3 Ui Dashboard Fidelity Openai Gpt Image 2 High 01

Screening grade: partial. High-01 dashboard hits all required OCR strings, four-card order, and three stacked zones on charcoal/cyan/off-white; only material miss is unspecified chart axis tick/time text. Rubric: artifact-severity=4/5, count-accuracy=5/5, layout-fidelity=5/5, ocr-exact-match=5/5, prompt-adherence=4/5, …

Size
1672×941
Seed
not exposed
Method
official api
View experiment →
Generated specimen r3-ui-dashboard-fidelity-openai-gpt-image-2-high-02 r3-ui-dashboard-fidelity-openai-gpt-image-2-high-02
successGPT Image 2 High

R3 Ui Dashboard Fidelity Openai Gpt Image 2 High 02

Screening grade: partial. High-02 matches title/status, exact four-card metrics, chart title, and three-zone structure; residual issue is extra chart axis numerals not listed in the frozen prompt text set. Rubric: artifact-severity=4/5, count-accuracy=5/5, layout-fidelity=5/5, ocr-exact-match=5/5, prompt-adherence=4/5, …

Size
1672×941
Seed
not exposed
Method
official api
View experiment →