GoldieBench Blog · 9 min read
Free AEO Audit: How To Get One, What It Can't Tell You, And Which Free Models Can Help
A free AEO audit shows if ChatGPT, Perplexity and Google's AI name you. The free routes, what they miss, and free models that score well on GoldieBench.

A free AEO audit is a no-cost check of whether AI tools like ChatGPT, Perplexity, Gemini and Google's AI Overviews name your brand, cite your pages and get your facts right.
My short answer to "is it worth it?" is yes, as long as you know what a free audit can't tell you.
The quickest way to get one with an expert is to book a free SEO strategy session with my team, which works like a free AEO audit conversation.
AEO stands for answer engine optimisation, which is the work of getting your brand named and cited in AI-written answers.
In this post I'll explain how a free AEO audit works, the free routes I checked today and where a free audit stops being enough.
Near the end, I'll show you which free and open models score well on this leaderboard, in case you want an AI to help you sort through the results.
What is a free AEO audit?
A free AEO audit is a first look at your AI visibility that costs you nothing but time.
It answers whether AI tools mention you for the questions your buyers ask, which competitors they name instead and which websites they cite.
A free audit isn't a piece of software with its own AI model, because the answers it records come from the AI tools themselves.
The free routes to an AEO audit
I checked each of these on its own page on 17 October 2026.
| Route | What its own page says | What to watch for |
|---|---|---|
| My free strategy session | A free AI brand visibility session with a custom visibility plan, an AI visibility roadmap and a competitor gap analysis, 100% free with no obligation | It's my own agency, so weigh my bias, and it's a call rather than a written report |
| HubSpot AI Search Grader | A free one-time check of what ChatGPT, Perplexity and Gemini say about you based on their training data, scored out of 100 | You complete a form to see full results, and training data may differ from live answers |
| Bing Webmaster Tools | Free reports, including an AI Performance report for citations in Copilot and Bing's AI summaries | It only covers Microsoft's AI experiences and select partners |
| Google Search Console | A free service whose Performance report counts AI Overviews and AI Mode traffic in the "Web" search type | Google reports it alongside your other web search clicks |
| Go Fish Digital AI Visibility Audit | A free, expert-led AI visibility audit with a live 30-minute strategy session | I haven't used it, so I'm describing its page only |
| A DIY prompt panel | You ask real buyer questions in each AI tool and log the answers | It takes the most time, but it shows live answers |
Microsoft introduced the AI Performance report in public preview on 10 February 2026.
How to run a free AEO audit yourself
This is the shortest version that still gives you a useful baseline.
1. Check the crawlers
Open your robots.txt file and make sure OAI-SearchBot, PerplexityBot and Googlebot aren't blocked on your key pages.
OpenAI says OAI-SearchBot surfaces websites in ChatGPT's search features, and Perplexity says PerplexityBot surfaces and links websites in its search results.
I read OpenAI's crawler page and Perplexity's crawler page on 17 October 2026.
2. Write ten buyer prompts
Write ten questions a buyer would ask before choosing a business like yours, mixing best-of, comparison, problem, brand and alternatives questions.
3. Ask them in three AI tools
Ask each one in a fresh chat in the three AI tools your buyers use most, and run each one twice if you have time.
4. Log what you see
Write down whether you were named, who else was named, which websites were cited and whether your facts were right.
5. Pick three fixes
Fix crawler blocks first, wrong facts second, and make the most-cited missing source your first coverage goal.
📞 Want an expert to read your free audit with you? Book a free SEO strategy session and my team will show you why competing brands appear more often and give you a clear AI visibility roadmap.
What a free AEO audit can't tell you
This is the part that decides whether a free audit is enough for you.
| Limit | Why it happens |
|---|---|
| It can't show what every buyer sees | AI answers change from one chat to the next, so any audit is a sample |
| It can't explain why a model chose a source | None of the companies behind ChatGPT, Perplexity, Gemini or Claude publishes that formula |
| It can't cover every phrasing | Ten or twenty prompts are a sample of the questions buyers ask |
| A grader score can't replace live answers | HubSpot's page says its grader works from the models' training data |
| It can't guarantee a future answer | Nobody can guarantee what a model says |
Google also says you don't need new machine-readable files, AI text files or special schema.org markup to appear in AI Overviews or AI Mode.
So if a free audit ends by selling you one of those as the fix, be careful.
When is a free AEO audit worth it, and when isn't it?
A free audit is worth it for almost everyone as a first step, because it costs nothing and gives you a dated baseline.
It's enough on its own when you have one main service, a short list of competitors and the time to re-run your prompts every month.
It stops being enough when you sell in several markets, when the competitor gap is large or when the fix is third-party coverage.
My strategy session page explains why coverage matters, because it says AI platforms don't recommend brands simply because those brands publish more on their own websites.
It says they look for consistent signals across the wider web, such as trusted publications, expert articles, independent reviews, comparisons and credible third-party references.
Earning that coverage takes relationships and time, and that's usually where people want help.
📞 Think your gap is bigger than a free audit can fix? Grab a free SEO strategy session with my team and we'll map the bylines, reviews, comparisons and research that fit your market.
Free AEO audit vs a full AEO audit
The free version is a first pass, and the full version goes wider on every axis.
| Part | Free AEO audit | Full AEO audit |
|---|---|---|
| Prompts | Ten buyer questions | 20 to 50 buyer questions across more types |
| AI tools | The three your buyers use most | ChatGPT, Perplexity, Gemini, Claude, Google AI Overviews and Google AI Mode |
| Runs | Once or twice per prompt | At least twice per prompt, re-run every month |
| Technical checks | A quick robots.txt look | Crawler, CDN, indexing, snippet and schema checks on every key page |
| Facts and coverage | Wrong facts and the most-cited missing source | A full fact file and a ranked list of coverage gaps |
| Output | A count of mentions and three fixes | A scored baseline and a prioritised fix list |
I'd always start free, because the first pass tells you whether the full version is worth your time.
How different businesses should use a free AEO audit
The method is the same for everyone, but the questions that matter change with the business.
A local service business should focus on "best [service] in [town]" questions, because those are the ones closest to a call.
A software company should focus on comparison and alternatives questions, because buyers use them to build a shortlist.
An online shop should focus on product facts, because a wrong price or spec in an AI answer puts buyers off.
A consultant should test their own name and "best [expert] for [problem]" questions, because bylines and reviews are the kind of third-party references my strategy session page says AI platforms look for.
An agency should run a free audit for every new client, because the dated log becomes the baseline for every report that follows.
How to keep a free AEO audit repeatable
A free audit is only useful if you can run it again in the same way next month.
Save your prompts word for word, because a small change in wording can change the answer.
Write the date on every row, because the trend matters more than any single result.
Keep a column for the AI tool and the mode you used, such as Google AI Overviews or Google AI Mode, because Google says they can use different models and techniques.
Copy the cited URLs in full, because the domains that keep appearing are your map for earning coverage.
Re-check your robots.txt file every month too, because a security or CDN change can block a crawler without anyone noticing.
Which free models can help you analyse a free AEO audit?
Now for the GoldieBench part, and I need to be clear about what it measures.
GoldieBench scores models on one-shot builds such as landing pages, apps, games and visuals, on a 0 to 10 scale.
It doesn't measure how well a model sorts an audit log, and it doesn't measure which brands an AI recommends.
A free AEO audit doesn't ship its own model, but you might want a model to help you tidy hundreds of logged answers or draft answer-first rewrites.
If you want that step to stay free, these are the free or open models on the board, next to the paid models behind the big assistants for reference.
| Model | Maker | Cost on its GoldieBench page | GoldieBench average |
|---|---|---|---|
| GLM-5.2 | Zhipu / Z.ai | Open weights, free for individuals | 7.77 |
| Qwable 5 27B Coder | DJLougen, on a Qwen base | Free, runs locally on MLX only | 7.14 |
| Qwen 3.7 | Alibaba | Open weights, free for individuals | 7.00 |
| Claude Opus 5 | Anthropic | Paid API | 8.27 |
| GPT-5.6 Sol | OpenAI | Paid API | 8.16 |
| Gemini 3.6 Flash | Paid API | 7.08 |
GLM-5.2 is the highest-scoring free option in that table at 7.77, which puts it above Gemini 3.6 Flash's 7.08.
Qwable 5 27B Coder averages 7.14 and runs on your own machine, which suits a private audit log you'd rather not upload.
You can see more models that run offline on the local models board.
The paid frontier models still score higher, with Claude Opus 5 at 8.27 and GPT-5.6 Sol at 8.16.
What the model scores mean for a free AEO audit
The scores don't change how you run the audit, because you always test the AI tools your buyers actually use.
They matter only for the helper step, where a model sorts your log or drafts rewrites for you.
For that step, the brief you give the model matters more than the model, because a clear fact file and a human check catch most mistakes.
A free local model with a good brief can be enough for sorting a spreadsheet of answers.
I'd still have a person approve every rewrite before it goes live, whichever model drafted it.
Related reading on GoldieBench
My sibling post on the full AEO audit explains every step and the models behind the answers.
My King of AEO post looks at answer engine optimisation from the model side.
In the same batch, I also wrote about the free AEO report, which shows how to write up what you find.
If you'd like my team to look at your results, you can book a free SEO strategy session and leave with a plan for traffic, mentions and leads.
Also On Our Network
- Agent Operatorsmy 30-minute DIY method for a free AEO audit
- AI Income Deskfour ways a business can get a free AEO audit
- AI Tool Verdictsix free AEO audit options rated side by side
- agentos.guidea free weekly AEO audit that runs inside your Agent OS
- agentos.guidehow my Agent OS runs answer engine optimisation
Run the free version first, know its limits and add expert help only where the gap is real, and a free AEO audit will be worth every minute.
FAQ
How do I get a free AEO audit?
You can book a free SEO strategy session with Julian Goldie's team, which works like a free AEO audit conversation, or run a DIY prompt panel. Free tools such as HubSpot's AI Search Grader, Bing Webmaster Tools and Google Search Console add more data.
Is a free AEO audit worth it?
Yes, as a first step, because it costs nothing and gives you a dated baseline of what AI tools say about you. It stops being enough when the competitor gap is large or when the fix needs third-party coverage.
What can't a free AEO audit tell me?
It can't show what every buyer sees, explain why a model chose a source, cover every phrasing or guarantee a future answer. A grader based on training data may also differ from live search answers.
Does a free AEO audit use its own AI model?
No. A free AEO audit records answers from the AI tools themselves, such as ChatGPT, Perplexity, Gemini and Google's AI features. It doesn't ship a model.
Which free models score best on GoldieBench?
Among free or open models, GLM-5.2 averages 7.77, Qwable 5 27B Coder 7.14 and Qwen 3.7 7.00. GoldieBench measures one-shot builds, not audit analysis or brand recommendations.
Can a free AEO audit guarantee a mention in ChatGPT?
No. Nobody can guarantee what a model says. A free audit shows you where you stand today and what to fix first.


