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 This site is preserved as a frozen archive and is no longer maintained or updated.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 →

Experiment · complete

Organic Chrome refinement: Balanced

Which model lane best preserves this isolated Organic Chrome fidelity direction?

Runs
9
Variable
provider/model lane
Prompt
oc26r-r4a-balanced-prompt v1.0

Protocol

Targeted continuation of the Organic Chrome campaign, isolating capsule-ended lamella geometry, fan-to-coil topology, emerald optical material, crop, and negative space.

Frozen prompt template

Square 1:1 abstract campaign image, no text or logo. Near-black green-black seamless field. One enormous asymmetrical procedural object occupies only the upper-right and right side: 48 to 72 separate smooth elongated glass lamellae, each a broad flattened capsule-ended ribbon rather than wire. They share a hidden origin beyond the top-right edge, fan leftward in dense parallel arcs, then the lower layers return into a cropped nested coil at bottom-right. The left 45 percent and lower-left remain almost empty darkness. Material is emerald-tinted optical resin with partial liquid-chrome reflection, transparent green body, dark green edges, mint and icy-cyan strip-light highlights, internal refraction, and crisp rounded surfaces. Extreme macro crop, structure exits top and right edges, no horizon or floor. Controlled ray-traced studio lighting, rich blacks, high contrast, slight analog scan softness. Preserve a clear two-level hierarchy: an upper fan of long comb-like blades and a lower overlapping helicoid return. Individual blades stay visibly separate and evenly repeated. Avoid a freestanding ring, wreath, basket, shell, flower, feather, literal plant, hair-thin wires, mesh, smoke, or extra object.

Review rubric

  • Prompt Adherence
  • Material Realism
  • Negative Space
  • Geometric Coherence
  • Campaign Readiness
  • Originality

Notes

3 unseeded repetitions per lane. Exact prompt is byte-identical within every provider/model comparison cell. External reference remains unpublished.

Evidence

Recorded runs

Generated specimen oc26r-r4a-balanced-grok-imagine-2-0-01 oc26r-r4a-balanced-grok-imagine-2-0-01
successGrok Imagine Image 2.0

Oc26r R4a Balanced Grok Imagine 2 0 01

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1024×1024
Seed
not exposed
Method
first party tool
View experiment →
Generated specimen oc26r-r4a-balanced-grok-imagine-2-0-02 oc26r-r4a-balanced-grok-imagine-2-0-02
successGrok Imagine Image 2.0

Oc26r R4a Balanced Grok Imagine 2 0 02

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1024×1024
Seed
not exposed
Method
first party tool
View experiment →
Generated specimen oc26r-r4a-balanced-grok-imagine-2-0-03 oc26r-r4a-balanced-grok-imagine-2-0-03
successGrok Imagine Image 2.0

Oc26r R4a Balanced Grok Imagine 2 0 03

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1024×1024
Seed
not exposed
Method
first party tool
View experiment →
Generated specimen oc26r-r4a-balanced-grok-imagine-quality-01 oc26r-r4a-balanced-grok-imagine-quality-01
successGrok Imagine Quality

Oc26r R4a Balanced Grok Imagine Quality 01

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1024×1024
Seed
not exposed
Method
first party tool
View experiment →
Generated specimen oc26r-r4a-balanced-grok-imagine-quality-02 oc26r-r4a-balanced-grok-imagine-quality-02
successGrok Imagine Quality

Oc26r R4a Balanced Grok Imagine Quality 02

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1024×1024
Seed
not exposed
Method
first party tool
View experiment →
Generated specimen oc26r-r4a-balanced-grok-imagine-quality-03 oc26r-r4a-balanced-grok-imagine-quality-03
successGrok Imagine Quality

Oc26r R4a Balanced Grok Imagine Quality 03

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1024×1024
Seed
not exposed
Method
first party tool
View experiment →
Generated specimen oc26r-r4a-balanced-openai-gpt-image-2-medium-01 oc26r-r4a-balanced-openai-gpt-image-2-medium-01
successGPT Image 2 Medium

Oc26r R4a Balanced Openai Gpt Image 2 Medium 01

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1254×1254
Seed
not exposed
Method
official api
View experiment →
Generated specimen oc26r-r4a-balanced-openai-gpt-image-2-medium-02 oc26r-r4a-balanced-openai-gpt-image-2-medium-02
successGPT Image 2 Medium

Oc26r R4a Balanced Openai Gpt Image 2 Medium 02

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1254×1254
Seed
not exposed
Method
official api
View experiment →
Generated specimen oc26r-r4a-balanced-openai-gpt-image-2-medium-03 oc26r-r4a-balanced-openai-gpt-image-2-medium-03
successGPT Image 2 Medium

Oc26r R4a Balanced Openai Gpt Image 2 Medium 03

Reviewed on the labeled refinement contact sheet against the unpublished operator reference. This run is retained as matrix evidence; model ranking is reported at campaign level.

Size
1254×1254
Seed
not exposed
Method
official api
View experiment →