ora
chatbase.com
success
Claude Code · Haiku 4.50:55
intent

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

11steps
55.6sduration
$0.5651cost
128,593tokens
16 steps5 reasoning steps2 searches
home
docs
/features
docs
docs
search
docs
/reviews/chatbase
search
home
docs
67%
on-site discovery
78%
reliability
22%
link following
path origin
  • previous resource22%
  • web search33%
  • prior knowledge44%
insight

The agent successfully completed a discovery-evaluation task by combining inaccessible direct site content with third-party review aggregators and comparison articles. Chatbase.com's homepage and pricing pages are heavily JavaScript-rendered, returning minimal static content; the agent worked around this by searching for and fetching independent reviews (SiteGPT, Chatimize) that contained the substantive information needed to explain what Chatbase does, who it's for, pricing tiers, and competitive positioning. The site itself publishes pricing on /pricing, but the agent could not extract it from the JavaScript-heavy response and instead relied on third-party summaries.

  • Steps [8] and [12] (web searches) surfaced review aggregators and comparison articles; steps [9] and [10] (SiteGPT and Chatimize reviews) were the actual content sources for pricing, features, and differentiators. The agent did not extract this information from Chatbase's own site.
  • Chatbase.com returns only HTML boilerplate and script tags in steps [1], [2], [13], [14]—no readable pricing tables, feature lists, or value propositions in static form. The /docs endpoint (step [6]) exists but was not followed up; the agent determined the site was 'heavily JavaScript-based' and pivoted to search.
  • The agent cited sources from Chatbase's own pricing page in the final response, but those citations appear to be drawn from prior knowledge or the third-party summaries (step [9] explicitly mentions credit-based pricing and model-dependent costs), not from a successful parse of Chatbase's own HTML.
  • Missing from the agent's research: direct inspection of the Chatbase app UI, the /docs knowledge base content, or live rendering of the JavaScript pages. The agent relied on external reviews to infer pricing structure, feature gaps, and competitive weaknesses.
  • The agent correctly identified three confusions: the opaque credit system, JavaScript-heavy marketing pages, and missing comparison tables—all observations grounded in what it could not extract from Chatbase.com itself.

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 →