682 AI-generated landing pages, judged

Same brief, different model, real output. Each page was generated as HTML, rendered on desktop and mobile, and graded by a vision judge. Nothing is mocked up — click any card to see the render and what the judge said.

See the full ranking on the Frontend & Landing Pages leaderboard.

Gemini 3.5 Flash vs Glm 5 (6 shared briefs, biggest score gap first)

All models Claude Fable 5 Claude Haiku 4.5 Claude Opus 4.5 Claude Opus 4.6 Claude Opus 4.8 Claude Sonnet 4.5 Claude Sonnet 4.6 Claude Sonnet 5 Deepseek v3.1 Terminus Deepseek v3.2 Deepseek v4 Pro Gemini 3 Flash Preview Gemini 3.1 Flash Lite Gemini 3.1 Pro Preview Gemini 3.5 Flash Glm 5 Glm 5.2 GPT-5 Mini GPT-5.4 GPT-5.4 Mini GPT-5.5 Grok 4.20 Grok 4.20 Beta Kimi k2.5 Kimi k2.6 Kimi k2.7 Code Minimax m2.7 Minimax m3 Mistral Medium 3.1 qwen3.5 Plus 02 15 qwen3.7 Max

Comparing best render per brief. Remove a model (×) to go back to the grid.

Glow & Grain — local bakery site (non-tech)

score gap 37

Glm 5 · max reasoning render for Glow & Grain — local bakery site (non-tech) — Glm 5 · max reasoning 35

Failed

Ferrovia — enterprise pricing tiers

score gap 6

Glm 5 · max reasoning render for Ferrovia — enterprise pricing tiers — Glm 5 · max reasoning 80

Strong · winner

Tradewinds — marketplace pricing

score gap 2

Glm 5 · max reasoning render for Tradewinds — marketplace pricing — Glm 5 · max reasoning 78

Usable · winner

Lumen Health — compliance-careful healthtech

score gap 1

Glm 5 · medium reasoning render for Lumen Health — compliance-careful healthtech — Glm 5 · medium reasoning 73

Usable

Cedar & Sage — DTC product page

score gap 0

Glm 5 render for Cedar & Sage — DTC product page — Glm 5 72

Usable

Northwind — B2B SaaS landing page

score gap 0

Glm 5 render for Northwind — B2B SaaS landing page — Glm 5 72

Usable

How these were made and judged

Each model received the same landing-page brief (a fictional company from our shared benchmark universe) and produced a single self-contained HTML page. We rendered every page at desktop and mobile sizes and had a vision judge grade the screenshots against the brief — hierarchy, copy, credibility, and conversion basics. Scores and judge quotes are unedited. Full methodology on the collection page.

This page is Spring Prompt, running

We just did this for every model. Do it for your prompt.

The rankings above come from running real tasks through real models and scoring every output. Spring Prompt is that same engine — pointed at your prompt, your test cases, and your definition of good.

Generate test cases from your prompt — no eval set required to start.
Compare models side by side with quality, cost and latency in one matrix.
Optimise the winner until the scores say it's ready to ship.

Join the waitlist Browse all benchmarks

Experiment · Cold outreach email

Prompt × model results

12 test cases · 3 evals

Claude Opus

GPT-5

Gemini

7.1

6.8

7.4

8.3

7.9

8.0

9.2 ★

8.6

8.4

Best combo: v3 × Claude Opus

9.2 quality · $0.004/run · 1.8s