e-Learning Formats in Moodle: The Big Overview for Education Providers
Anyone who runs Moodle is really managing a whole zoo of e-learning formats: SCORM packages from content providers, H5P exercises from your own editorial team, classic Moodle activities like Lesson and Page – and in between PDFs, videos and external tools. Every format has its strengths, its limits and its own way of storing learning content. And that is exactly what becomes the decisive question in 2026: because anyone using AI tutors quickly notices that the answers are only as good as the knowledge base the tutor can access – and that knowledge sits in precisely these formats.
This article is the kickoff of our series on e-learning formats in Moodle. Here you get the overview: what does each format do, how well is it integrated into Moodle – and how does its knowledge get into the vector database (RAG) of an AI tutor? A dedicated detail article follows for each format.
Key Takeaways
Moodle is a patchwork of formats – and that's okay: SCORM, H5P, xAPI and native activities all have their place. It only becomes a problem when the knowledge stays locked inside the formats.
Interactivity is not the same as AI readiness: A SCORM package can be highly interactive and still be a black box for the AI. What matters is whether the content can be extracted and synchronized into a knowledge base.
Tracking standards are not knowledge access: SCORM and xAPI govern what the LMS learns about the learner – not what an AI tutor knows about the content. Two different problems.
The path to a RAG knowledge base runs through sync: With Alphalearn's Moodle Sync, course content is automatically extracted, transferred into a vector database and linked to the AI tutor per course – including content from SCORM, xAPI and HTML packages.
Why e-learning formats determine the success of your AI tutor
A modern AI tutor answers questions not from the general world knowledge of the language model, but from a controlled knowledge base: a vector database that is searched via Retrieval-Augmented Generation (RAG). For this to work, the learning content has to get out of its formats and into this knowledge base – ideally automatically and kept up to date.
And this is where the formats part ways: A Moodle Page is open text and trivial to capture. A SCORM package is a ZIP container with its own logic whose contents have to be extracted. An H5P element stores its texts in a JSON structure. Anyone planning their AI tutor should therefore ask both questions: how good is the format for learners – and how readable is it for the AI?
The most important e-learning formats at a glance
SCORM
the industry standard for exchangeable courses
SCORM (Sharable Content Object Reference Model) has been the standard for exchanging finished e-learning modules between authoring tools (Articulate, Adobe Captivate, iSpring …) and LMSs for over 20 years. Moodle fully supports SCORM 1.2, and SCORM 2004 with limitations (no sequencing).
Strengths: reusability, completion and score tracking, huge content ecosystem.
Weaknesses: the package is a closed box – Moodle only sees course completion and exercise results, not the content.
AI and SCORM: Alphalearn by Alphabees can ingest SCORM content into the knowledge base and even embed the AI tutor directly into the SCORM package – turning the static module into an interactive learning environment.
xAPI & cmi5
learning data beyond the LMS
xAPI (Tin Can API) solves SCORM's biggest weakness: it tracks learning experiences everywhere – in apps, simulations, even offline – and stores them as Actor-Verb-Object statements in a Learning Record Store (LRS). In Moodle this requires plugins (e.g. Logstore xAPI) plus an external LRS; cmi5, the designated SCORM successor profile, is also connected via plugin. Strengths: granular learning analytics, device independence. Weaknesses: infrastructure effort, and like SCORM, xAPI delivers tracking – not content access.
AI and xAPI: xAPI statements reveal what learners have done – valuable for analytics and learner profiles. The content itself enters the knowledge base via package extraction; Alphalearn supports integration into xAPI packages natively.
H5P
interactive content directly in Moodle
H5P is the most popular format for self-created interactive content in the German-speaking region: interactive videos, drag-and-drop exercises, presentations, quizzes – created directly in the browser, part of the core since Moodle 3.9 with its own Content Bank. Strengths: no authoring-tool costs, open technology (HTML5/JSON), xAPI statements included. Weaknesses: tracking less standardized than SCORM, content quickly spreads across many individual elements.
AI and H5P: H5P stores texts structured in JSON – easily extractable. Via Alphalearn's Moodle Sync, the content of H5P activities automatically lands in the knowledge base, and the AI tutor can answer questions about exactly the exercise a learner is currently stuck on.
Moodle activities
Page, Book, Lesson, Quiz & Co.
The native Moodle activities are the backbone of most courses: Page for individual content pages, Book for multi-page chapters, Lesson for branching learning paths, Quiz for exams, plus Glossary, Forum and file resources (PDF, video). Strengths: zero additional cost, maintained directly in the LMS, cleanly structured. Weaknesses: limited interactivity compared to H5P or authoring-tool content.
AI for Moodle activities: For RAG, native activities are the most rewarding format: open, structured text. Alphalearn's automatic course sync captures them without detours – including scheduled updates when course content changes.
Formats compared:
| Format | Interactivity | Tracking | Moodle integration | RAG knowledge base (AI tutor) |
|---|---|---|---|---|
| SCORM | high (depending on authoring tool) | completion & score | Core (1.2 full, 2004 limited) | ⚠️ extraction needed – possible via Alphalearn Sync |
| xAPI / cmi5 | high (tool-dependent) | granular (LRS required) | via plugin + LRS | ⚠️ extraction needed – possible via Alphalearn Sync |
| H5P | very high | xAPI statements, gradebook | Core (since 3.9, Content Bank) | ✅ JSON structure, easily extractable |
| Moodle activities | medium | Moodle built-in | native | ✅ ideal – open text |
| PDF / Video / Files | low | access | as a resource | ✅ PDF directly, video via transcript |
As of July 2026. RAG suitability rates how well content can be automatically transferred into an AI knowledge base.
From format to knowledge base:
How Moodle content becomes an AI tutor
The path is the same for all formats – only the effort differs:
Extraction: The learning content is read out of its format – open text directly, SCORM/xAPI packages and H5P elements via structured extraction, videos via transcripts.
Vectorization: The content is split into sections and stored as vectors in a database – the RAG knowledge base.
Linking: Each AI tutor is linked to the knowledge base of its course, so answers are course-specific and verifiable.
Synchronization: When course content changes, the knowledge base updates itself – without manual upkeep.
Conclusion & outlook on the series
There is no single right e-learning format – a good Moodle course usually combines several. What is decisive for 2026 is that the content does not stay locked inside its formats but becomes usable as a knowledge base for interactive, AI-supported learning support. In the coming parts of this series we examine each format individually: SCORM, H5P, xAPI & cmi5 as well as the Moodle activities – each with concrete instructions on how to get its knowledge into your AI tutor.
Frequently Asked Questions
-
Moodle natively supports SCORM (1.2 fully, 2004 with limitations), H5P (part of the core since Moodle 3.9 with the Content Bank) as well as its own activities like Page, Book, Lesson and Quiz. xAPI and cmi5 are connected via plugins and an external Learning Record Store.
-
SCORM tracks learning progress within the LMS (completion, score). xAPI tracks learning experiences everywhere – including outside the LMS – and stores them granularly in a Learning Record Store. Both govern tracking, not access to the content itself.
-
Not on its own – a SCORM package is a closed box. Solutions like Alphalearn extract the content into a RAG knowledge base and can additionally embed the AI tutor directly into the SCORM package.
-
Via an automatic course sync: content is extracted, split into sections, stored as vectors and linked to the course’s AI tutor. With Alphalearn this runs on a schedule and automatically – including for external sources like Google Drive or SharePoint.
-
Different problem: H5P is a format for creating interactive content directly in Moodle, SCORM an exchange standard for finished modules from authoring tools. Many courses use both. For the AI knowledge base, H5P is easier to capture thanks to its open JSON structure.
-
Currently there is no AI tutor that can read videos directly. However, using a combination of visual-understanding models and a transcription tool, it is possible to bring video content into a text form and upload it to the knowledge base of AI tutors.