Chat & Companion · Reviewed 2026-05-23

SiteGPT

STEADY · 72/100

Competent website-trained chatbot SaaS in a crowded category — does what it says, but doesn't out-position Chatbase, Custom GPTs, or Intercom Fin on anything obvious.

Visit SiteGPT →

SiteGPT is the kind of product that emerged confidently in 2023 and now has to defend its ground. The core function — train a chatbot on your site's content, embed it as a widget, escalate to human on demand — is solid, and the marketing site is unusually candid for the category (a working live demo on the homepage, transparent published pricing, no 'book a call' wall on basic plans). What it lacks is a sharp axis of differentiation. The 'ChatGPT for your products' framing is shared by at least a dozen rivals: Chatbase, Cody, MyAskAI, OpenAI's Custom GPTs, and incumbents Intercom Fin / Zendesk AI / Drift. Pricing is competitive but not the cheapest; features are complete but not exceptional. The presence of an API and webhook integrations is a meaningful credit for agent-driven workflows — but the API docs are gated behind a sign-up, which limits how much an evaluating agent (like Hlido's tester or yours) can verify in advance. For a buyer with no preference between rivals, SiteGPT is a defensible default; for a buyer with specific needs (multilingual, voice, deep CRM tie-in), most competitors have a story SiteGPT doesn't yet emphasise.

Why STEADY

STEADY (72) because the product works, ships on time, and shows enough operational maturity (working demo, transparent pricing, escalate-to-human) to be defensible. Not VITAL because the category is saturated with functionally equivalent options and SiteGPT doesn't have an obvious 'use this over X' axis. Not FADING because there's no evidence of decay — just middle-of-the-pack positioning.

What it does well

What it fails at

Red flags

Best for

  • SMBs / mid-market buyers who want a turnkey AI support widget without overthinking the choice
  • Teams that need 'good enough' faster than they need 'best in class'
  • Use cases where escalate-to-human matters more than answer-quality at the margin
  • Builders embedding chatbots into custom workflows via API (the API existing is a real plus here)

Not recommended for

  • Enterprises with strict data-residency / compliance needs (SiteGPT's public data posture is too thin)
  • Multilingual-first use cases (specialised rivals like Botpress invest more here)
  • Voice-first support (out of scope)
  • Buyers who care which underlying LLM serves their users (SiteGPT doesn't disclose on public pages)
  • Anyone wanting to A/B test the product before signup beyond the demo widget

Compared to

Agent relevance

API Webhook

API + webhooks are advertised, which makes SiteGPT addressable by agent-driven workflows (e.g., pre-screen tickets, trigger on conversation events). Docs gated behind signup, so an evaluating agent must register to confirm capabilities. No MCP server.

Agent-friendly score: 5/10

Evidence

Public-surface checklist

scorecard.json · registry · methodology

Verdict by Hlido Editor · Method: public-surface-tier-1+editorial-narrative-v2 · Methodology version 2026.05 · Next review due 2026-08-23