How it works
AI models¶
remake asks 12 AI models, sold as 3 model sets: Standard, Pro and Hyper. Each schedule asks one set, and every prompt goes to every AI model in it.
The 12 AI models¶
| AI model | First in | Asked on a new schedule |
|---|---|---|
| ChatGPT | Standard | Yes |
| Gemini | Standard | Yes |
| Meta AI | Standard | Yes |
| Google AI Overviews | Standard | Yes |
| Claude | Pro | No |
| Perplexity | Pro | No |
| Qwen | Pro | No |
| Mistral | Pro | No |
| Google AI Mode | Hyper | No |
| Grok | Hyper | No |
| Kimi | Hyper | No |
| DeepSeek | Hyper | No |
Some screens shorten two names to AI Overviews and AI Mode.
Copilot is not tracked for now. Its earlier answers stay in your history.
The model sets¶
Each set includes the one before it. Pro and Hyper take a saving off the AI models.
| Model set | AI models | Saving |
|---|---|---|
| Standard | The first 4 | None |
| Pro | Standard's 4, plus Claude, Perplexity, Qwen and Mistral | 10% off |
| Hyper | Pro's 8, plus Google AI Mode, Grok, Kimi and DeepSeek | 20% off |
A run costs your active prompts on every AI model in the set, plus a fixed part for the website steps. The fixed part is the same in every set. Prices and limits has worked totals for each set.
Each set's price is in the Rates list in your wallet, once you're signed in. Every run shows its total before it starts.
Pick a model set¶
You pick the set in two places:
- Setup, step 3: under Which AI models we ask. See Start your first run.
- Budget & Spending: in the Models step of the schedule dialog. See Schedules.
Each card shows the number of AI models, the saving and the price of one run with your prompts. You can't add or drop single AI models; a schedule always asks its whole set.
Changing the set changes the next run. Past runs keep the AI models they asked.
How answers are collected¶
Each prompt is asked live, on the day of the run, the way a customer would ask it. The answer is recorded as it comes back, word for word, with the links it shows.
Each AI model is reached one of 3 ways:
| How it's asked | AI models |
|---|---|
| Read from the real product: the answer a customer sees on screen, from a fresh chat with no history | ChatGPT, Gemini, Google AI Overviews, Google AI Mode |
| Asked with web search on: the model reads today's web pages for the prompt, then answers | Claude, Grok, Kimi, DeepSeek, Mistral, Qwen, Meta AI |
| Asked with its own search | Perplexity |
No AI model is asked from memory alone. A model without live search misses many businesses a customer's own search would find. How we measure explains why.
Meta AI is Meta's own AI model, asked with web search on. It is not the meta.ai app. That app answers only people signed in to a Meta account, so no one can read it for you.
The models asked with web search on read the same pages for a given prompt, so their sources often overlap. The wording and the businesses each one names are its own.
The 4 models read from the real product take a minute or two per answer, so their answers land last in a run.
ChatGPT and Gemini have been read from the real product since 8 October 2026. Before that they were asked with web search on, so a trend line for either can step at that date.
Each prompt is asked once per AI model per run. If an AI model can't answer, that answer is refunded and left out of your numbers.
A ChatGPT or Gemini answer that comes back cut off is asked again once. If it is still cut off, it is refunded and left out.
A few runs ask differently:
| Run | AI models |
|---|---|
| The free run after email confirmation | ChatGPT with live search, 3 prompts |
| A pitch | ChatGPT, Perplexity and Google AI Overviews |
| A bulk run | Perplexity |
Countries and languages¶
| Setting | What it changes | What it doesn't change |
|---|---|---|
| Country | Every prompt belongs to your brand's country. The AI writes place words for it, such as "in Switzerland". | Where the AI models are asked from |
| Location | A city written into the prompt's words, such as "phone shop in London" | Where the AI models are asked from |
| Language | The AI writes your prompts in English, whatever your country. Prompts you add yourself can be in any language. | Which AI models answer |
ChatGPT, Gemini, Google AI Overviews and Google AI Mode are asked from the United States. The other AI models search the web with no location set.
To track a city, put it in the prompt. See Prompts, topics and locations.
Where to find it¶
- Overview: Visibility by AI model shows how often each one names you.
- Insights: the AI model filter, and the By AI model chart.
- Prompts: one cell per AI model for each prompt. With 6 or more AI models, that needs the All AI models view; Summary folds them into one count.
- Rates: the Rates dialog, from Rates → on the Wallet, lists every AI model's price.
FAQ¶
Can I ask only Claude?¶
No. A schedule asks a whole model set. Claude comes with Pro and Hyper.
Does Hyper give me more features?¶
No. Standard, Pro and Hyper are sets of AI models for a schedule. Every account gets every feature, and each brand picks its own set.
Why does an AI model show a dash?¶
It gave fewer than 3 answers in that view, too few for a number. Widen the time window or clear a filter.