The monthly cost of tracking a brand in ChatGPT, Gemini and other AI answers is prompts × engines × countries × runs × credits per answer. Credits per engine, what each plan covers, and four budgets from 480 to 600,000 answers a month.
The formula
Every answer is one task: one prompt, sent to one engine, in one country. A tracker sends the same prompts on a schedule, so five numbers set the monthly bill: how many prompts you track, on how many engines, in how many countries, how often, and what each engine's answer costs in credits.
answers per month = prompts × engines × countries × runs per month
credits per month = prompts × countries × runs per month × (credits of every engine, added up)In this guide a month holds 4 weekly runs or 30 daily runs. Each engine's price in credits is on the pricing page, and a monitor charges the same credits per answer as a single request, so the formula covers scheduled runs and one-off collection alike.
Credits per answer by engine
Engines come in two price steps. Answers from ChatGPT, Perplexity, Bing Copilot and Meta AI cost 1 credit. Answers from Gemini, Qwen, Doubao, Baidu Wenxin, Baidu AI Search, Google AI Overviews, Google AI Mode, Naver AI Briefing and Naver AI Tab cost 1.25 credits. The table turns those credits into dollars per 1,000 answers on the smallest and the largest plan.
| Engine | Credits per answer | Per 1,000 answers on Basic | Per 1,000 answers on Max |
|---|---|---|---|
| ChatGPT | 1 | $2.50 | $1.00 |
| Perplexity | 1 | $2.50 | $1.00 |
| Bing Copilot | 1 | $2.50 | $1.00 |
| Meta AI | 1 | $2.50 | $1.00 |
| Gemini | 1.25 | $3.125 | $1.25 |
| Qwen | 1.25 | $3.125 | $1.25 |
| Doubao | 1.25 | $3.125 | $1.25 |
| Baidu Wenxin | 1.25 | $3.125 | $1.25 |
| Baidu AI Search | 1.25 | $3.125 | $1.25 |
| Google AI Overviews | 1.25 | $3.125 | $1.25 |
| Google AI Mode | 1.25 | $3.125 | $1.25 |
| Naver AI Briefing | 1.25 | $3.125 | $1.25 |
| Naver AI Tab | 1.25 | $3.125 | $1.25 |
Two options add to an answer. include.html stores the HTML of the answer page for 0.25 credits on ChatGPT, Perplexity, Meta AI and Bing Copilot, charged when the page is stored, and Gemini, Google AI Overviews and Google AI Mode include it at no extra credit. include.analysis finds your brand, your competitors and citations of your own sites in the answer for 1 credit, charged when the analysis finishes. Routine tracking runs work well with both options off.
What each plan covers
Plans differ in the price of a credit. A larger plan lowers the price of every answer, and unused credits expire at renewal, so the right plan is the smallest one that covers a month of runs.
| Plan | Price per month | Credits | Per 1,000 credits | ChatGPT answers | Gemini answers |
|---|---|---|---|---|---|
| Basic | $20 | 8,000 | $2.50 | 8,000 | 6,400 |
| Pro | $200 | 170,000 | $1.18 | 170,000 | 136,000 |
| Scale | $500 | 480,000 | $1.04 | 480,000 | 384,000 |
| Max | $1,000 | 1,000,000 | $1.00 | 1,000,000 | 800,000 |
Four worked budgets
Each budget below tracks four engines: ChatGPT, Perplexity, Gemini and Google AI Overviews. Together they cost 4.5 credits per prompt, per country, per run. Swap in the engines your buyers use and the formula gives the new total.
| Setup | Answers per month | Credits per month | Plan that covers it |
|---|---|---|---|
| One brand, 30 prompts, 1 country, weekly | 480 | 540 | Basic, $20 |
| One brand, 50 prompts, 1 country, daily | 6,000 | 6,750 | Basic, $20 |
| Agency, 20 brands × 25 prompts, 1 country, daily | 60,000 | 67,500 | Pro, $200 |
| Platform, 1,000 prompts, 5 countries, daily | 600,000 | 675,000 | Max, $1,000 |
The weekly brand setup costs 540 credits a month, and the free credits cover its first 3 runs. A daily check of 50 prompts on ChatGPT alone is 1,500 answers and 1,500 credits, well inside Basic.
How many answers a reliable number needs
A mention rate is a share of answers, so its precision depends on how many answers stand behind it. A rate from 30 answers carries a margin of about 18 points either way, 100 answers bring it to about 10 points and 400 to about 5. How many AI answers you need works through the math.
Turn the sample you need into a schedule. To see a 10-point change between two weeks on one engine, each week needs about 400 answers from that engine: 60 prompts run daily give 420. To see a 20-point change, about 100 answers per week are enough: 15 prompts run daily, or 100 prompts run weekly.
Ways to keep the bill small
- Run weekly in steady categories and daily around launches and campaigns. A weekly schedule costs about a seventh of a daily one.
- Add a country where you sell or plan to sell. Every country multiplies the bill.
- Pick engines by where your buyers ask. ChatGPT and Perplexity answers cost 1 credit, Gemini and Google AI Overviews answers 1.25.
- Spread answers over more prompts and fewer repeats. 30 prompts asked once say more about a category than 3 prompts asked ten times.
- Turn on HTML and analysis for the runs you review by hand.
Frequently asked questions
How much does it cost to track one brand in ChatGPT?
50 prompts checked daily on ChatGPT are 1,500 answers and 1,500 credits a month. The Basic plan at $20 a month covers it with 8,000 credits.
What does one AI answer cost?
One ChatGPT answer costs 1 credit: $0.0025 on Basic and $0.001 on Max. One Gemini or Google AI Overviews answer costs 1.25 credits: $0.003125 on Basic and $0.00125 on Max.
How many free credits does a new account get?
500 credits, enough for 400 to 500 answers depending on the engine.
Which plan fits an agency tracking many brands?
20 brands with 25 prompts each on four engines, checked daily, use 67,500 credits a month, which the Pro plan covers. The formula scales to any number of brands: add up the prompts across brands and multiply.
How often should I run my prompts?
Weekly suits most brands and reads the trend at a seventh of the daily cost. Daily suits launches, campaigns and fast-moving categories, and gives enough answers to see a 10-point change week over week with about 60 prompts.
