The short answer
No. Whizi has no public or developer API, no API key issuance, and no personal access tokens. Every route on the Whizi backend authenticates as an end user session rather than as a third-party integration, so there is no credential to create, paste, rotate or revoke. This is true on every plan, including Powerhouse.
The upstream provider credentials Whizi uses to reach the models are server side secrets on the backend and never leave it.
| What people ask for | Status in Whizi |
|---|---|
| A REST or streaming chat API | Does not exist |
| An API key or secret you can generate | Does not exist |
| A personal access token | Does not exist |
| A credential that lets an outside tool act on your account | Does not exist |
| A team, organization or workspace with seats | Does not exist |
| A full export of your own account data | Exists, as Export Data in Settings, guest sessions included |
If you need to call models from your own code, Whizi is not the tool for that job. Whizi is a chat workspace: one subscription, one picker, a 280+ model catalogue. The section below covers what is supported for getting data in and out.
What the terms of service say about scripting it
The absence of an API is not an oversight waiting for a workaround. Driving the product with a script is prohibited in the terms, in two separate places.
Section 3, user representations, has you warrant that "you will not access the Services through automated or non-human means, whether through a bot, script or otherwise".
Section 8, prohibited activities, lists it again from the other direction. You agree not to "Engage in any automated use of the system, such as using scripts to send comments or messages, or using any data mining, robots, or similar data gathering and extraction tools." The same section separately forbids you to "Systematically retrieve data or other content from the Services to create or compile, directly or indirectly, a collection, compilation, database, or directory without written permission from us", and, standard search engine and browser usage aside, to "use, launch, develop, or distribute any automated system, including without limitation, any spider, robot, cheat utility, scraper, or offline reader that accesses the Services".
The errors a script hits
Because the backend only knows end user sessions, anything automated runs into session and rate limit refusals rather than an integration error. A caller with no session token, or an expired one, gets the ordinary 401 unauthorized refusals catalogued in login and account access. A caller that does authenticate as a browser session is then refused by the ordinary per user rate limits, which are in rate limits and what they mean.
One refusal is specific to scripts. More than 3 device or session keys in 24 hours is the guard behind HTTP 403 account_sharing_limit, so a script that starts a fresh session per run looks exactly like a shared account and gets refused on that basis. It clears on its own once the window rolls forward.
What to use instead
Getting all your data out. Export Data sits in Settings under Account and returns the whole account as one downloadable JSON file, guest sessions included. The mechanics of that file are in exporting your chats and files.
Getting data in. Chat import accepts exactly two sources, ChatGPT and Claude. The export file never reaches the server: the browser reads conversations.json or the raw export .zip, parses it, and posts normalized conversations. Every imported conversation is keyed by account plus source plus source id, so re-importing the same export or clicking Import twice lands the same chats once. Import Chats sits in Settings under Account.
Publishing an answer somewhere else. A conversation can be published behind an unguessable public link that is read-only, stripped of user ids, emails and internal fields, and revocable at any time. See share a conversation.
Getting structured output rather than prose. A chat answer can produce a real .xlsx or .docx in your browser rather than a payload, which is documented in exporting your chats and files.
Reusable instructions across conversations. There is no account-level custom instructions setting, but a project carries its own instructions, up to 32,000 characters, applied to every conversation in it, plus up to 10 pinned files that ride along as prompt text on every turn. That covers the "same system prompt every time" use case without a script.
The question that usually arrives alongside this one is whether Whizi has team or workspace accounts. It does not, and that has its own page: team and workspace plans.
- No public or developer API exists, on any plan
- No API keys and no personal access tokens are issued
- Every backend route authenticates as an end user session
- Provider credentials are server side secrets and never leave the backend
- Scripted and automated access is prohibited in sections 3 and 8 of the terms
- A script hits session, sharing and rate limit refusals, not an integration error
- Account data export, chat import, share links and generated files are the supported routes in and out
- There are no team, organization or workspace accounts either
Frequently asked questions
Is there a Whizi API key on any plan?
No. No tier issues one. Starter, Pro and Powerhouse differ in what the plan opens and how much of it you get, and none of them adds a credential, an integration surface or a developer console. The tier differences themselves are in plans and limits.
Can I use my Whizi subscription in my own app or script?
No. There is no endpoint to call and no credential to call it with, and the terms of service prohibit automated access in two separate sections. Practically, a script would authenticate as a browser session and then be refused by the account sharing guard once more than 3 device or session keys are seen in 24 hours, or by the 30 requests per minute all-routes ceiling.
Whizi calls model providers, so can I get at those keys?
No. The upstream provider credentials Whizi uses to reach the models are server side secrets on the backend and never leave it.
Can I schedule the data export to run on its own?
There is nothing to call on a schedule. The export is a button in Settings under Account rather than a route you can hold a credential for, and the terms prohibit driving the product with a script. What the file contains is in exporting your chats and files.
Could a third-party tool connect to my Whizi account?
Not through any credential you can issue. Nothing in Whizi lets an outside tool hold delegated access to your account, so anything claiming to connect would be driving a browser session, which is what the terms prohibit and what the account sharing guard refuses.
Is a developer API on the roadmap?
As of August 2026 there is no public API, no API key issuance and no personal access token issuance in the product, and the terms treat automated use as prohibited rather than unbuilt.