saas · Aug 21, 2026
50/100 Thin
Scores above 48% of 1,252 multi-step funnels we have measured. Among the 959 saas funnels in that group, above 37%.
Coverage is detection coverage, not a verdict — it counts the conversion tactics we could prove are present, out of the 11 that apply to a saas funnel.
Measured by Roentgen 1.25.1, one of our versioned rulesets. A capture is read once and its measurements freeze there, so the stamp says which ruleset read this one.
Walk 18 links seen, 13 walkable, 1 matched, and we walked 2 pages.
the entry page’s own title, headline, buttons and prices — quoted, never paraphrased
measuredextracted directly from the live page. a fact, not an opinionLLM Gateway for Coding Agents | OmniaKey
An LLM gateway for coding agents. Use one OpenAI-compatible API key for Claude, GPT, Gemini, and Grok.
One LLM Gateway for Coding Agents
Change the base URL — keep the same API key
LLM Gateway for Coding Agents
Route leading coding models through one gateway
No subscriptions, no tiers, just usage-based pricing
About billing, models, and integration
LLM Gateway for Coding Agents Leading frontier models
SETUP
CONTROL+3 more, not shown
OpenAI
Anthropic
Gemini
COPY
cURL
Python
JavaScript
Show all 28 models →
Cookie Preferences
$1.2
$5
$6
$25
$0.12
$0.5
$0.35
$2.1+17 more, not shown
structural bands identified on this funnel, cropped from its own capture
measuredextracted directly from the live page. a fact, not an opinion
conversion tactics we could prove are present, with the on-page text that evidences each
measuredextracted directly from the live page. a fact, not an opinionGemini 3.1 Pro Preview 1M ctx
free credit
input, email, input, password
docs link
Why are selected models up to 93% cheaper than the official price?
free credit
applicable tactics we checked for and did not see on any step we reached
measuredextracted directly from the live page. a fact, not an opinionAbsence is a measurement, not a recommendation: we do not know why a tactic is missing, whether it belongs here, or whether adding it would change anything. Where a walk stopped early, a tactic living deeper would also read as not detected.
contrast, mobile, typography, accessibility, weight — each with what would fix it
measuredextracted directly from the live page. a fact, not an opinion59% of measurable text passes WCAG AA. 4+ runs fail. 25% of text was unmeasurable (over images/gradients). not counted either way.
worst: “01” at 2.11:1 (22px, #b8b2a5 on #ffffff). needs 4.5:1
fix Raise text/background contrast to ≥4.5:1 (≥3:1 for large text), especially on CTAs and body copy.
Responsive viewport meta is present.
Smallest body text is 11px (base ~18px). below the ~16px readable floor.
fix Keep body copy ≥15-16px; reserve sub-13px for fine print only.
1 paragraph runs past ~95 characters per line.
fix Cap measure at ~60-75ch for comfortable reading.
html[lang] is set for screen readers.
3 headings skip a level (e.g. <h2> followed by <h4>). Screen-reader users navigate by this outline.
fix Step heading levels one at a time; use CSS for size, not a deeper level.
At least 0.3 MB across 64 measurable resources, but 33 cross-origin resources did not expose a size. The true weight cannot be determined from the page.
4 stylesheets/scripts in <head> block first paint.
fix Defer non-critical JS (defer/async) and inline only critical CSS.
15 elements look clickable but cannot be reached by keyboard.
“01”, “02”, “03”
fix Use <button>/<a>, or add role + tabindex="0" and a key handler.
4 fields carry validation attributes (required/pattern/aria-invalid).
palette by on-page usage · type rendered live
measuredextracted directly from the live page. a fact, not an opinionBody copy in their text role, on a surface card.
Body copy set at their measured paragraph size. This is how their product reads to a first-time visitor. same family, same scale, same weight.
We also looked for Elements and Playbook match, and found none on this funnel.
