Design and build Fjora's responsive Taste the North campaign landing page from the frozen prompt, rules, facts, photography, and requirements.
An exact model and reasoning setting per Configuration and no subagents, but no shared sandbox: Codex used its own, and Claude Code and OpenCode had none until 2026-09-24. Runs are compared only within one protocol. How the protocols differ
One archived screenshot per published Run against version 1.2.0, of the Run’s representative Revision. Ordered by setup and then by repetition, not by quality.
Frozen with this version. Requirement IDs are stable across versions and are cited unchanged by evaluation records.
Automated checks are not an accessibility certification. Recorded axe findings describe what the tooling observed at capture time, on a pinned browser and viewport. A check that reports no violation is not proof that the interface is accessible, and a selector problem is inconclusive rather than an implementation failure.
| # | Model | Tool | Revision | Status | Actions |
|---|---|---|---|---|---|
| Fig. 01 | Claude Fable 5.1 (low) | Claude Code | 0 · initial | live | |
| Fig. 02 | Claude Opus 5 (high) | Claude Code | 0 · initial | live | |
| Fig. 03 | Claude Opus 5.5 (medium) | Claude Code | 0 · initial | live | |
| Fig. 04 | Claude Sonnet 5 (high) | Claude Code | 0 · initial | live | |
| Fig. 05 | GPT-5.6-Luna (max) | Codex CLI | 0 · initial | live | |
| Fig. 06 | GPT-5.6-Sol (max) | Codex CLI | 0 · initial | live | |
| Fig. 07 | GPT-5.6-Sol (low) | Codex CLI | 0 · initial | live | |
| Fig. 08 | GPT-5.6-Terra (max) | Codex CLI | 0 · initial | live | |
| Fig. 09 | GPT-5.6-Terra (low) | Codex CLI | 0 · initial | live | |
| Fig. 10 | GPT-6-Astra (max) | Codex CLI | 0 · initial | live | |
| Fig. 11 | GPT-6-Astra (low) | Codex CLI | 0 · initial | live | |
| Fig. 12 | GPT-6-Astra (high) | Codex CLI | 0 · initial | live |
Fig. 01Claude Fable 5.1 (low)
Fig. 02Claude Opus 5 (high)
Operator-attested: the Configuration, tool version, displayed model, and elapsed time were recorded by the operator from what the tool showed at Run time. They are not independently verified. An exact model ID appears only when the tool exposed one; it is never inferred.
Published sites are independent, AI-generated demos served from separate preview origins with local mock data. They are not products, and no UIBenchmark session data crosses to them.
Exact model ID, tool version, elapsed time, run state, stop reason, discarded attempts and every published Revision, repairs included, in run log order. Each Run and each Revision links to its own page.
01Run 1 · Claude Fable 5.1 (low) · Claude Code
These Runs were executed against version 1.2.0, the most recently frozen Challenge Version under protocol v2. Runs against another frozen Version are listed on that Version’s own page, because they received different inputs.
Design and build Fjora's responsive Taste the North campaign landing page from the frozen prompt, rules, facts, photography, and requirements.
Fictional Nordic botanical sparkling-water campaign
The workflow designs the interface itself from the frozen brief.
Read PROMPT.md and BENCHMARK_RULES.mdThis is the brief of the most recently frozen Challenge Version under protocol v2. The full text of its launch prompt and shared rules, and the frozen inputs of earlier versions, are on the Challenge Version pages, so nothing here is a paraphrase of what a Run actually received.
The declared rules and viewports version 1.2.0 was executed and captured under.
Fig. 03Claude Opus 5.5 (medium)
Fig. 04Claude Sonnet 5 (high)
Fig. 05GPT-5.6-Luna (max)
Fig. 06GPT-5.6-Sol (max)
Fig. 07GPT-5.6-Sol (low)
Fig. 08GPT-5.6-Terra (max)
Fig. 09GPT-5.6-Terra (low)
Fig. 10GPT-6-Astra (max)
Fig. 11GPT-6-Astra (low)
Fig. 12GPT-6-Astra (high)
02Run 1 · Claude Opus 5 (high) · Claude Code
03Run 1 · Claude Opus 5.5 (medium) · Claude Code
04Run 1 · Claude Sonnet 5 (high) · Claude Code
05Run 1 · GPT-5.6-Luna (max) · Codex CLI
06Run 1 · GPT-5.6-Sol (max) · Codex CLI
07Run 1 · GPT-5.6-Sol (low) · Codex CLI
08Run 1 · GPT-5.6-Terra (max) · Codex CLI
09Run 1 · GPT-5.6-Terra (low) · Codex CLI
10Run 1 · GPT-6-Astra (max) · Codex CLI
11Run 1 · GPT-6-Astra (low) · Codex CLI
12Run 1 · GPT-6-Astra (high) · Codex CLI
Discarded attempts failed for a technical reason — a crashed session, a tool outage, a broken upload — and were discarded and retried from the same frozen inputs. The count is not a quality signal.
Generation cost: Not separately metered. Runs were executed on the operator’s personal tool subscriptions, which do not itemize a per-Run charge. No amount is attributed, and no zero-cost claim is made.
Starter and asset hashes are published so every run can be shown to have received the same bytes. There is no starter download and no asset download.
Starter SHA-256
c719934a49c705dd266e4cfa55c3b68faa36fc0f8f095aaee848cb0678dfb68d
Asset pack SHA-256
4ba7a63f9ec62eaf342e775e9f8ce984967ac5433999555e48f9960599c242bb