Trust & security

Honest billing

• Usage is billed at the model provider’s actual token cost plus a 25% platform fee — the same rule for every model.
• Every reply shows what it cost, and every charge is listed in your usage history.
• If you stop a reply early, you pay only for what was generated.
• Cancel in two clicks in Settings → Plan & billing; we ask once if you’d like to keep your plan; access continues until your period ends.
• We email you before every renewal. See the Refund Policy.

Honest model labels

• Every reply is labelled with the model that actually answered it.
• If a model has to be swapped — for a regional restriction or an outage — the chat says so.
• When Auto chooses a model for you, the reply shows which one it chose.
• Generated images carry a visible “AI-generated” label and an embedded provenance record.

More in our AI Disclosure.

Your data

• We don’t sell your personal information, and we don’t use your chats to train AI models.
• People in the EU and UK are never routed to providers we exclude for those regions.
• You can download all your data as a file from Settings at any time.
• You can delete your account yourself; it is removed after a 30-day grace period you can cancel.
• Every company that processes your data is listed on our subprocessors page.

Security

• The site and app are served over HTTPS only.
• Most people sign in with a one-time email code or Google. Passwords, if you set one, are stored only as a salted one-way hash.
• You can see every device signed in to your account, sign out any of them, and we email you when a new device signs in.
• Card and UPI payments are handled by Razorpay; your full card details never reach our servers.

Report a security issue

Found a vulnerability? Please tell us through our contact form (choose “Bug”) before sharing it publicly, and give us a reasonable time to fix it. Our security.txt has the same details. Service health is on our status page.