How to Track Your Brand's AI Visibility for Free in 2026
Paid AI-visibility trackers cost $69 to $2,500+ a month. Here's the free manual method, plus a free open-source self-hosted tool, for tracking how ChatGPT and Claude talk about your brand.
Contents5
Why "ranking on Google" stopped being the whole story
A stat worth sitting with: one recent industry analysis found that 73% of brands with zero mentions in AI chatbot answers were still ranking on page one of Google. Traditional SEO and AI visibility are turning out to be two different games, and most of the tooling for the second one is expensive. The paid AI-visibility trackers on the market right now run from roughly $69 a month on the low end to $2,500 and up for enterprise platforms.
That's a real gap if you're a solo founder, a small agency, or anyone who wants to know how ChatGPT, Claude, Gemini, and Perplexity talk about their brand without signing up for a platform built for marketing teams with budgets.
The manual method, and why it's more work than it sounds
The lowest-cost approach costs nothing but time: write down the 15 to 20 prompts a real customer would type when they're close to a buying decision, your "money prompts," and run each one manually in ChatGPT, Claude, Gemini, and Perplexity once a day for a couple of weeks. Log whether you show up, where, and what the AI actually says about you versus your competitors.
It works, but it doesn't scale. Four assistants times 20 prompts times daily checks is 80 manual queries a day if you're thorough, and AI answers vary between runs even for the identical prompt, so a single check on a single day tells you less than you'd think.
A free, self-hosted option for AI visibility tracking
This is the gap [Lettertrace](https://questloops.com/tools/lettertrace) is built for. It's an open-source, MIT-licensed tool that automates the same idea, generating realistic prompt variations from a topic, running them on a schedule against Claude, ChatGPT, Gemini, and Google AI Overviews, and tracking visibility, sentiment, and share of voice against named competitors over time. The catch that makes it free: you bring your own API keys from Anthropic, OpenAI, and Google, so you pay each provider directly for the queries instead of paying a markup to a SaaS platform.
| Manual tracking | Lettertrace (self-hosted) | Typical paid platform | |
|---|---|---|---|
| Cost | Free (your time) | Free (your API usage) | $69 to $2,500+/month |
| Models covered | Whichever you check by hand | Claude, ChatGPT, Gemini, Google AI Overviews | Varies by vendor |
| Scheduling | Manual, whenever you remember | Daily or weekly, automatic | Automatic |
| Competitor benchmarking | Manual notes | Built in | Built in |
| Setup effort | None | Self-hosting + API keys | Account signup |
What actually moves the needle once you're tracking it
Knowing your AI visibility score is only useful if you know what changes it. A few factors show up consistently across research on the topic:
- Pages that lead with a direct, one-to-two sentence answer before the supporting detail get cited far more than pages that build up to the point.
- AI systems tend to cite independent, third-party sources (reviews, comparison sites, Wikipedia, press coverage) noticeably more than they cite a company's own marketing pages.
- Organization and product schema markup correlates with higher citation rates, since it gives the crawler an unambiguous entity to attach facts to.
- Citation behavior differs sharply by platform. Perplexity cites far more often than ChatGPT does per response, so a brand that looks invisible in ChatGPT logs might still be showing up regularly in Perplexity.
Getting started this week
Pick your money prompts first, before you pick a tool. A tracker only tells you something useful if you're asking it the questions your actual customers ask. Start with five to ten prompts covering your product category, your top two or three competitors by name, and one or two "best X for Y" style questions, then decide whether the manual method covers your needs or it's worth the half hour of setup to self-host Lettertrace with your own keys.
If you're also trying to get more out of what you're already paying for on the model side, [our breakdown of cutting AI agent API costs](https://questloops.com/blog/how-to-cut-your-ai-agent-s-api-bill-token-compression-smart-routing-and-when-a-gateway-pays-for-itself) covers the token-spend side of the same bring-your-own-key approach.