A quick walkthrough of the Playground, safety layer, and billing — plus answers to the questions parents and educators ask most often.
Our safety layer is built on the Understory framework: it isn't enough to just refuse harmful content. A child-safe AI has to do three things at once.
Block method disclosure and harmful content — graded on the C-SSRS crisis scale.
Route to trusted adults and crisis resources. Resist parasocial dependence.
Scaffold thinking with Socratic questions instead of doing the child's work for them.
In Settings → Profile, pick the age band that matches your child (8–11, 12–15, 16–18, or adult). CanopyKids uses this to shape response tone, reading level, and the Socratic scaffolding it applies during learning conversations.
Open the home page and type a question. You can pick a model (Claude, GPT, or Gemini) from the composer. New accounts get 10 free messages so you can try it before subscribing.
For learning questions, CanopyKids won't just hand over the answer — it asks age-appropriate guiding questions first. Kids can still get the answer, but only after the model has helped them think it through.
Every message is inspected for crisis signals, self-harm content, and unhealthy parasocial patterns. When something serious surfaces, CanopyKids routes to trusted adults and safety resources — this can't be turned off.
The dashboard shows messages used, remaining quota, and the current billing period. Parents on a family plan can also see a weekly Parent Digest summarizing themes and any safety events.
Head to Pricing to pick a plan or switch tiers — upgrades and downgrades are prorated automatically. Use “Manage billing” to update your card, download invoices, or cancel (access continues until the period ends).
Email anu@middymind.ai and we'll get back to you within one business day. For urgent safety concerns involving a child, contact local emergency services first.