ora
koyeb.com
success
Claude Code
intent

What does koyeb.com do and who is it for? Explain it back to me.

13steps
57.2sduration
$0.1716cost
89,047tokens
16 steps3 reasoning steps5 searches
home
docs
/features
docs
search
docs
search
search
search
docs
docs
docs
search
62%
on-site discovery
88%
reliability
25%
link following
path origin
  • previous resource25%
  • web search38%
  • prior knowledge38%
insight

The agent successfully assembled a comprehensive overview of Koyeb by combining prior knowledge with fetches from the site's homepage, pricing page, docs, and comparison pages. The site's JavaScript-heavy architecture forced reliance on web search to surface key pages, and pricing details required navigation through multiple doc sections rather than a single, machine-readable source. The agent worked around scattered feature information and incomplete homepage content to deliver a coherent answer.

  • ›Steps [1], [2], [5] returned heavily JavaScript-rendered pages where critical information (pricing tiers, features, free-tier status) was embedded in the DOM rather than exposed in initial HTML, limiting direct parsing.
  • ›Steps [10], [11], [12] (comparison pages and pricing FAQ) were explicitly cited in the final response's Sources section and contained the most concrete details: comparison tables vs. Heroku/Render, GPU pricing ($2.50/hr for H100), bandwidth costs ($0.04/GB), and the free-tier deprecation after Mistral acquisition.
  • ›The site lacked a dedicated, easily discoverable 'Features' page—the agent had to infer features from comparisons, blog posts (discovered via web search), and scattered docs. The homepage positioning did not clearly surface GPU/AI capabilities or global deployment reach in a fetchable form.
  • ›Web searches [6], [8], [9], [13], [14] acted as routers to specific docs (e.g., pricing FAQ, comparisons) but the search snippets themselves did not contain answer content; they were navigation aids. The agent relied heavily on prior knowledge (38% sourced) to frame the overall narrative.
  • ›Pricing model required synthesis from three sources: the pricing page (HTML didn't render detail), the pricing FAQ doc, and external third-party pricing aggregators (srvrlss.io, spotsaas.com) linked in search results. No single page on koyeb.com provided a clear, complete pricing table.
  • ›The Mistral AI acquisition (February 2026) and free-tier deprecation were mentioned in external sources but not prominently communicated on the main site—a significant positioning gap for a major platform change.

Want to run your own?

Join the waitlist for early access to point your own agents at any domain, with the intents you choose.

or talk to us about agent readiness →