Guide ยท Decision-Stage
Can you learn AI SEO for free?
15 of 44 AI SEO programs cost nothing and cover as much syllabus as paid ones: what free teaches, what it skips, when paying helps. Checked September 2026.
Yes, for most of the syllabus. 15 of 44 English-language programs in this directory cost nothing (4 could not be verified; checked 18 September 2026), and those free programs confirm a mean of 7.3 curriculum topics against 7.1 for paid programs with a published price. Free is not a thinner curriculum. What you give up is structure and a person who will look at your pages and tell you what is broken. If you want the ranked view before reading further, the free AI SEO course hub orders them by score.
What does free material reliably cover?
13 of 15 free programs teach LLM citations and 12 teach AI visibility measurement, with 1 unverified on the citations flag (checked 18 September 2026). Citations and measurement are the strong end of the free tier. That makes sense: both are teachable from public screenshots and a spreadsheet, and both are what people arrive asking about.
| Topic | Free programs that teach it | Not verified |
|---|---|---|
| LLM citations | 13 of 15 | 1 |
| AI visibility measurement | 12 of 15 | 1 |
| GEO | 9 of 15 | 2 |
| AEO | 6 of 15 | 2 |
| AI crawler access | 7 of 15 | 1 |
| Local AI SEO | 5 of 15 | 1 |
All figures checked 18 September 2026. Read the "not verified" column as "the public page does not say", not as "no". Checks cover what a provider publishes where a buyer can see it before handing over an email address.
Where does free material thin out?
5 of 15 free programs cover local AI SEO, 1 unverified (checked 18 September 2026), so a single-location business looking for map pack and assistant answers together will run out of material quickly. Local work is the clearest gap. Answer engine optimisation and AI crawler access both sit near half.
The second gap is format. 4 of 44 English-language programs are a published manual or guide with no lessons attached. A manual is fast to read and gives you nothing to hand a junior on Monday. Where video hours are published at all, the median across the directory is 2.6 over 17 programs, with 27 not publishing a runtime.
Currency is harder to check than coverage. 18 of 44 English-language programs put dates on their reviews or testimonials, so a module recorded before the current assistant features shipped can sit behind a page that looks new. Look for a dated update log, or a lesson that names an engine behaviour you recognise from recent months.
Why do vendor academies teach around their own tool?
13 of 44 English-language programs run on a vendor's own academy platform, 0 unverified (checked 18 September 2026). These are typically free and typically shaped by the product: the measurement lesson walks through that vendor's dashboard, the crawler lesson matches what that vendor's crawler reports, and the workflow lesson assumes you have a seat.
None of that is hidden, and none of it makes the teaching wrong. It does mean the method you learn may not survive a switch of tooling.
One practical test: open the measurement module and count how many steps you could repeat with a spreadsheet and manual prompting. If the answer is none, you are learning a product, not a technique.
Free access is usually traded for contact details, and sometimes for a referral. 15 of 44 programs run an affiliate scheme, with 28 unverified. Worth knowing when a free lesson recommends a paid tool by name.
What does free give up in public evidence?
4.1 of six transparency checks is the mean for free programs, against 4.4 of seven for paid ones over 29 programs (checked 18 September 2026). The refund check does not apply when nothing is charged, but instructor identity, dated updates and a published syllabus all still do.
Across the whole English-language set, 36 of 44 programs name an instructor with a public track record, and 8 name nobody at all.
Evidence is thinner too. 24 of 44 English-language programs sit at low confidence and are listed without a rank everywhere they appear, because there is not enough public review material to score them fairly. Free programs are over-represented in that group: people who paid nothing rarely leave a review. The scoring rule behind that is set out in how the scores are calculated.
One thing money does not buy is accreditation. Of the programs that issue a certificate, 0 of 27 name an accrediting body, with 11 unverified. A paid certificate and a free completion badge carry the same external weight, which is none.
What does free not replace?
15 of 44 programs include a community and 15 of 44 run live calls or sessions, with 20 unverified on that check (checked 18 September 2026). Community-first formats account for 8 of 44 programs. A free manual cannot do what those do, which is answer a question about the page you are actually working on this week.
What is a sensible free-first order of work?
15 of 44 English-language programs are free, 4 unverified (checked 18 September 2026), which is enough choice to run a month of free material before spending anything. The table below is the order that wastes the least time, and the last row is the whole point: the gaps you write down become your actual buying brief.
| Step | What to check before you start |
|---|---|
| Pick the free program | Does the public page list a syllabus and a named instructor, or only a signup form? |
| Do the AI citation module first | Can you run the prompts yourself without a paid tool seat? |
| Build a measurement sheet | Track brand mentions across assistants weekly, by hand if needed |
| Ship one page change | Structured data or extraction formatting on a page you control |
| Note what you could not do | Those gaps, written down, are your actual buying brief |
Work through the free material for a month, keep a list of the moments you got stuck, and you will know whether the block is knowledge or feedback. If it is knowledge, another free program probably covers it. If it is feedback, you need a cohort or a community with people who answer questions.
When the free path does stall, the median normalised monthly cost across paid programs with a published price is $33.25 over 25 programs, and the full spread is in what AI SEO courses cost. Every English-language program sits in the comparison table in score order, free entries included.
So: after a month of free material, which task on your own site can you still not finish?
Frequently asked questions
Is free material enough to do AI SEO work for clients?
For the reading, yes. Free programs confirm a mean of 7.3 curriculum topics against 7.1 for paid programs with a published price (checked 18 September 2026). What free rarely includes is someone reviewing your own site and telling you what is wrong with it.
Do free courses cover GEO and AEO, or only basics?
Partly. 9 of 15 free programs teach GEO and 6 of 15 teach AEO, with 2 unverified. LLM citations are far more common at 13 of 15 (checked 18 September 2026).
Are vendor academy courses really free?
Usually free to enter, often behind an email form. 13 of 44 English-language programs sit on a vendor's own academy platform, and the lessons follow that vendor's product. Fine if you already use the tool, less portable if you do not.
Why are so many free programs listed without a rank?
Because public evidence is thin. 24 of 44 English-language programs sit at low confidence and are listed but not ranked on every hub. The rule is in [how the scores are calculated](/methodology/).