Ask anything.
Each question goes to the model best suited to it, with your personal details removed first — and every answer shows you what was decided.
The difference
The same question, asked two ways.
Choosing a model, judging whether a cheaper one would have sufficed, checking what personal information a prompt contains: today these are decisions you make yourself, if they are made at all. Governor makes them for you, on every question.
You make every decision
Decided before the question is sent
Works with the applications you already use. Any tool that talks to an OpenAI-compatible service — Open WebUI, LibreChat, LangChain and others — can be pointed at Governor once and is governed from then on.
To see what your AI assistant has already saved about you, paste its memory file into the scrubber →
How it holds up
Three promises, and why they hold.
Nothing to switch on, and nothing to take on trust: each can be verified from inside the product.
It chooses on evidence, not marketing
Models are ranked by how people rated their answers, one kind of task at a time — not by their makers' claims. Every answer states which model produced it.
The everyday models are very good
The models included at no extra cost rate just behind the most expensive ones, at a small fraction of the price. Most questions are answered equally well by one of them, which leaves the premium models for the questions that need them.
The provider never learns who is asking
Your question reaches the model from AI Governor, not from your account: no name, no identifier, nothing to build a history under. Because different questions go to different models, what any one provider sees is a scatter of anonymous questions, not a person. Conversations stay in your browser and can be cleared from Settings in one click.
The classification method, the learning process and the measurements behind them are set out in the paper.
Pricing
Unlimited, because most questions do not need an expensive model.
Start free in the browser, with no account. Paid plans remove the daily cap entirely: unlimited use of the everyday models — the best of which rates just behind the top of the leaderboard — plus a monthly balance for the premium models, spent only when a question warrants it. The desktop version is free, with no cap, because it runs on your own machine.
Free
$0 forever
Ten questions a day in the browser, or unlimited use on your own computer.
- 10 free questions a day in the browser
- The first five need no account
- Removes personal details from every prompt
- Selects the best-suited model for each question
- Learns your vocabulary over time
- 300+ models to choose from
- Or download it and run it without limits on your own machine
First 100 members
Founding
$8 /month, fixed for life
Unlimited everyday models, at a price that never increases.
- Everything in Free
- Unlimited use of the everyday models
- $1/month toward premium models
- Fully capable from the first day — no learning period
- Immediate access, no trial, and a price fixed for as long as you remain subscribed
AI Governor Pro
$18 /month
The standard plan, with a seven-day trial before any charge.
- Everything in Founding
- Unlimited use of the everyday models
- $1/month toward premium models
- Free for seven days; cancel at any time before the trial ends
AI Governor Pro+
$68 /month
For those who use the strongest models most days.
- Everything in Pro
- $75/month toward premium models — more than the plan itself costs
- Approximately 120 conversations on the strongest models
- Balance resets on your renewal date
AI Governor Max
$200 /month
Frontier models as the default rather than the exception.
- Everything in Pro+
- $225/month toward premium models
- Approximately 360 conversations on the strongest models
- Team plans from $149/month — contact us
Your premium balance goes further here than in any product where you choose the model yourself. Given a menu, most people select the largest model for everything; Governor reads the question first and uses an expensive model only when the task requires it. There is no API key to configure and no separate provider bill — every reply is included, on every plan, including the free one.
Questions
The questions people ask first.
Do I have to stop using ChatGPT or Claude?
No, and you do not need them to use AI Governor. The browser version includes the models: Claude, GPT, Gemini, Qwen and others are selected for you question by question, so there is nothing to connect and no other subscription is required. The free desktop application runs on your own machine and can send everything to a local model instead. Whatever you continue to use elsewhere is unaffected.
Can you see my prompts?
In the desktop application, everything happens on your computer and your prompt never reaches us. In the browser, we relay it, because that is how it reaches the model that answers — but we retain nothing: no prompt logs, no conversation store, nothing on our side to leak, subpoena or sell. Conversations are held in your browser only, which also means they do not follow you to another device, and clearing your browser data removes them. The provider receives a request from AI Governor, not from you: no account, no address, nothing to file it under. If a message is sensitive enough that you would rather the words themselves were changed, redaction is a single setting.
Will answers be worse without my context?
Your full question still reaches the model that answers it. What Governor keeps private is the separate step of determining what kind of question it is, which needs only a few words. Because a better-suited model is chosen, answers often improve.
Which model answers, and where does it run?
Whichever suits the question. The roster covers models from Anthropic, OpenAI, Google, Alibaba, xAI, Mistral and others, and each answer names the one used. Most everyday questions are answered by a Qwen or Gemini model, because they rate near the top at a fraction of the price, which means most questions are answered in Singapore or the United States; the full list is published rather than left to inference. When a question needs current information, the model searches the web itself and the sources appear beneath the answer. If you would prefer that nothing is sent anywhere, the desktop application can be pointed at a model on your own machine and everything is routed there.
What does it cost?
The desktop application is free and runs on your own machine. In the browser, the replies are paid for by us: there is no API key to configure and no separate provider bill. Paid plans provide unlimited use of the everyday models plus a monthly allowance for the premium ones; when that allowance is used up nothing stops — routing narrows to the everyday models and the conversation continues.