Daily scans
Every active prompt runs on your daily engines once a day, at the time you added it, and on Claude once a week.
Platform
See how AI answers your buyers’ questions
Add the prompts that matter to your market. Citeflare asks every engine each day and records whether your brand was named, where in the answer, and how it was described.
How it works
How prompt tracking works
01
Add your brand and prompts
Enter your brand name, aliases and domain, then add the questions buyers ask. Citeflare can also suggest prompts from your site and market.
02
Pick your engines
Choose the engines your plan includes. Each prompt runs on every selected engine, so you compare like with like.
03
Scans run every day
Each prompt is sent to the five core engines daily and to Claude weekly. The full answer and its sources are stored, so you can always read what was said.
04
Read the result
See visibility, position and sentiment per prompt and per engine, and how they move over time.
Capabilities
Everything in one place
6 engines, 10 on Enterprise
ChatGPT, Google AI Overviews, Google AI Mode, Gemini, Perplexity and Claude side by side; Grok, DeepSeek, Mistral and Llama on Enterprise.
Mention position
Where your brand appears in the answer: named first, third, or buried in a list.
Sentiment
Each mention is classified as positive, neutral or negative, with the snippet that shows why.
Brand aliases
Add alternative names and spellings so “Acme”, “Acme Inc” and “acme.com” count as one brand.
Full answer history
Open any scan to read the exact answer and the sources the engine used.
In depth
What you can do with it
Engine × prompt
One table, every engine
Answers differ by engine, so averages hide the story. The prompt table shows each engine in its own column and lets you spot where you are named and where you are missing.
- Filter by topic, engine or date range
- Click a cell to read the answer behind it
- Group prompts by topic to see themes
Trend
Know if the work is moving the needle
Visibility is plotted over time, so a content push or a site change can be judged against a baseline instead of a hunch. The final day is dashed because it is still filling in.
- Visibility, share of voice and position as tabs
- Compare periods
- Per-country tracking per project
Sentiment
Read how you are described, not just if
Being named is half the picture. Each mention carries a sentiment label and the sentence it came from, so you can catch an engine repeating an outdated or negative description.
- Positive, neutral and negative counts
- The exact snippet for every mention
- Spot answers that mis-describe your product
Metrics
What we measure, defined Every number has one definition, so you can check it against the data.
- Visibility score
- The percentage of completed scans in the period where your brand is mentioned in the answer.
- Average position
- The mean position of your mentions among every brand the answer names, tracked or not, where 1 means you were named first. Lower is better.
- Sentiment
- The count of your mentions classified as positive, neutral or negative.
- Scan coverage
- Completed scans out of scheduled scans, so you can see when an engine failed to answer.
Related
Works with the rest of the platform
Citations
See which domains and pages AI engines cite, split into owned, competitor, editorial and user-generated sources, including Reddit and YouTube.
Competitors
Measure how often AI engines name you versus competitors. Share of voice, a heatmap by engine and competitor suggestions pulled from real answers.
Agent
Ask Citeflare’s agent about your visibility, citations and competitors in plain English. Run skills like a weekly recap and get answers grounded in your own scan data.
Faq
Questions, answered
Which AI engines do you scan?
ChatGPT, Google AI Overviews, Google AI Mode, Gemini and Perplexity daily, and Claude weekly. Starter covers three of the five daily engines; Growth and Agency cover all five. Grok, DeepSeek, Mistral and Llama are on Enterprise.
How often are prompts scanned?
Once a day for every active prompt. Pause a prompt and it stops running.
Are these real user conversations?
No. Citeflare sends your prompt to each engine and records the answer, the way a person would ask it. Results can vary between runs, which is why daily trends are more useful than a single answer.
Can I track different countries or languages?
Each project has its own country and language, so create a project per market. Agency allows up to 10 projects.
What if an engine does not name my brand?
The scan is recorded as not mentioned, and the full answer and its sources are kept so you can see who was named instead.
See what AI says about you. Your first scan runs the day you sign up.
Monthly or annual plans. Cancel any time.