SCORM vs xAPI vs cmi5 vs LTI: Which E-Learning Standards to Build Into Your Product

SCORM, xAPI, cmi5 and LTI do different jobs, so you rarely pick only one. Use SCORM or cmi5 to deliver courses, xAPI with a learning record store to track detail, and LTI 1.3 to connect external tools. Versions of LTI before 1.3 lost official support on June 30, 2022, so they are off the table for a new build.

You are building or growing an e-learning product, and every enterprise customer asks the same thing. Do you support our standard? Each answer adds scope, and the wrong one can lock you out of integrations, reporting or the next market. Teams ask which elearning standard to use long before they ask who has to build it. This guide is written for the Product Owner who has to decide what lands on the roadmap. It covers what each standard does, who builds which part and how to decide for your type of product, plus a short plan for adding standards in the right order so the product stays simple as it grows.

What are e-learning standards and why do they matter for your product and learning management system?

E-learning standards are shared technical rules that let learning content, platforms and tools support data exchange and work together across different systems and learning platforms. They decide how content is packaged, launched and tracked, so your choice shapes your integrations and your reporting. A learning management system that follows a standard can accept elearning content it has never seen before, and content that follows a standard runs on different LMS platforms, other LMS systems and learning environments. Learning management systems, authoring tools and external tools all depend on the same shared rules.

Standards turn into real decisions the moment you start LMS software development for your own product. To put it plainly, four of these elearning standards matter today. SCORM packages and tracks courses, xAPI records learning activity anywhere, cmi5 launches content from an LMS using xAPI, and LTI connects outside tools to a platform. A fifth, AICC, began in 1988 as a set of rules for computer based training in aviation, and the group behind it was disbanded in 2014, so it is legacy. It later added an HTTP communication protocol, and some legacy systems still speak it, but new projects start elsewhere.

Used well, standards give your product four things.

• Compatibility, so content and tools work with any compatible LMS without custom work for each customer.
• Reuse, so one piece of training content serves many courses and contexts.
• Consistent tracking, so completion data and learner progress arrive in a consistent format from every source.
• Centralized reporting, so data from multiple systems lands in one view and your reporting capabilities grow with the product.

The key point is that these standards are roles, not rivals. A modern platform combines several of them, and each one gets a single clear job. xAPI became an official international standard in 2023, which makes it a safe base for new tracking work. In a modern elearning environment the standards also decide how fast you can add the next customer, tool or report. A custom software development company can map which of the four your architecture needs before any code is written.

How do SCORM, xAPI, cmi5 and LTI differ in what they track and who has to build what?

SCORM packages courses and reports completion and scores, xAPI records learning activity anywhere in a learning record store, cmi5 launches LMS content using xAPI, and LTI connects external tools to a platform. They solve four different problems, so the real question is which problems your product has.

SCORM stands for Sharable Content Object Reference Model. Its content aggregation model wraps scorm content in a ZIP file with a manifest file that describes the course structure. The LMS plays the content in the browser and receives the results through a communication protocol that every compliant package follows, so the same package works on different LMS platforms. SCORM 1.2 came out in 2001, and SCORM 2004 added sequencing and fuller interaction data. SCORM places the burden of complexity on the LMS, which is why content authors like it and platform teams carry the work.

xAPI records experiences as statements built from an actor, a verb and an object, such as a learner completing a module. Each time a learner interacts with an activity, the system can send a statement to a learning record store, so you can track learner activities from a mobile app, a simulation or real world activities, not only from a course. xAPI defines how to record learning data, not how to launch a course.

cmi5 is a profile of xAPI built for content launched from an LMS. Its first edition, called Quartz, dates from June 1, 2016. The LMS launches a unit of content, hands it a token and receives xAPI statements back. Nine defined verbs such as launched, completed, passed and terminated keep content delivery consistent across systems. The LMS must implement a full learning record store, so cmi5 moves work onto the platform side.

LTI, short for Learning Tools Interoperability, connects other systems and different tools to a platform without a separate login. Version 1.3 uses OAuth 2 and JSON web tokens for security, and its Advantage services add grade return, content selection through deep linking and access to user roles. LTI does not package content or track activity in detail, and it is the natural route for third-party educational content such as quizzes, labs and video. It solves integration, which is why it sits next to the other three rather than replacing them.

That's where it gets tricky, because the table hides the most useful split, which is who does the work. With SCORM the platform builds the runtime and the content author only exports a package. With xAPI the platform needs a record store, and every content source must send statements. With cmi5 the platform builds the launch logic and the record store. With LTI both sides build, because a platform acts as the host and each external tool acts as the guest. In custom LMS development the platform side carries the runtime, the record store and the launch logic, and the front end that plays and launches content is work for a React development company or an in-house team. If audit trails drive the decision, the xAPI vs SCORM question looks different for compliance training.

Is SCORM still relevant for a new learning product?

Yes, SCORM still has the widest LMS support, so a product that imports customer elearning courses needs it. The same goes for a product that runs next to a traditional LMS. It tracks mainly completion, pass or fail, score and time spent, so it is a weak base for detailed analytics. SCORM 1.2 stores only identifiers for interactions, while SCORM 2004 stores the full text of questions and answers and gives authors more control over how learners navigate and how course content rolls up. Neither version covers mobile learning, offline learning or activity outside the LMS. SCORM 1.2 remains the most widely adopted version across the elearning community, which is why years of learning materials still ship in it. Keep SCORM for corporate training libraries and the content your customers already own, and do not build advanced tracking on top of it.

Do you need a learning record store for the Experience API?

For xAPI and cmi5 you need a learning record store, because that is where statements are kept and read. SCORM and LTI do not need one. A learning record store receives statements from any source and can share data with other record stores. A statement is a short record of a learner interaction such as learner completed module, with an optional result and context. That structure is why xAPI can track offline learning and mobile learning, which matters for learners who work without a stable internet connection. It also lets systems exchange data and access data in a common format, so reports are not locked inside one product, and it captures more of the learning process than a completion flag.

The cost is real. You maintain the store, and you must agree on a consistent data structure early, because messy statements produce messy reports. If your roadmap includes detailed reporting or AI features that need structured data, plan the store from the start. If it does not, SCORM and LTI work without one.

Try our developers.
Free for 2 weeks.

No risk. Just results. Get a feel for our process, speed, and quality — work with our developers for a trial sprint and see why global companies choose Selleo.

Test our team

Share this offer

Which e-learning standard to use for your type of product?

Choose by job, not by popularity. Use SCORM or cmi5 to deliver courses, xAPI to track beyond the course and LTI 1.3 to connect tools, and combine them when your product has more than one job. Without LTI 1.3 every tool integration becomes a separate project, and adding xAPI late forces changes across the whole system. That holds for online courses and for training materials alike.

If your platform imports courses that customers made elsewhere, support SCORM first, because it must run on existing LMS platforms. It is the format your buyers expect, because their authoring tools export it. The authoring tools for e-learning your customers use decide which export formats you must accept, so ask early in every sales conversation which elearning tools they use and how they manage training content today.

If you ship your own courses in a new product, choose cmi5 instead of SCORM. It keeps the course structure and launch rules, and it records learning in xAPI format, so you do not outgrow it. Add SCORM import only when customers bring their own packages. A SaaS development company will ask you which customers import content and which connect tools, and that answer sets your first release.

If your product hosts third-party quizzes, labs, video or content libraries, build LTI 1.3 with its Advantage services, so one integration pattern covers multiple tools. Reject LTI 1.1, because official support for versions 1.0 to 1.2 and 2.0 ended on June 30, 2022. For growing EdTech scaleups the order of standards matters as much as the list.

If your roadmap includes analytics, personalization or AI features, add xAPI from the first release. Here's the thing. AI needs structured data to work and to deliver personalized learning experiences, and xAPI captures detailed learning events beyond course completion. Adding it later changes data models, content and reporting together, so it costs more than starting with it.

Think in three stages, because the right standard depends on where the product is today. At the MVP stage, ship the one standard your first paying customers require, which is SCORM import for a content platform or a single LTI integration for a tool platform. At the growth stage, add an event model with xAPI and a learning record store, so reporting and personalization have data to work with. At the enterprise stage, move new courses to cmi5, support LTI 1.3 with its Advantage services and add a reporting layer that your customers can trust. Each stage builds on the previous one, so nothing you ship early gets thrown away.

Use AICC only when a client asks for it, since it is legacy. Never treat the standards as interchangeable, and never force one standard to do two jobs. One workable combination is LTI for tools, xAPI for activity data and cmi5 or SCORM for structured courses.

How do you add e-learning standards without overbuilding your learning platform?

Start from how the product will be used, give each standard one clear role, define your data structure early and limit custom integrations. Add a standard when a real use case needs it, not because it exists. Adding standards after launch raises cost, and an inconsistent data structure is a known reason implementations fail. Follow these five steps.

1. Define how the system will be used, meaning who will create content, who launches it and what data the business needs back.
2. Assign each standard a clear role, such as SCORM or cmi5 for courses, xAPI for activity data and LTI for tools, so no two overlap.
3. Define the data structure early, including which verbs and activities you record, because consistent statements make reports and AI features possible.
4. Limit custom system integrations, and replace each with a standard wherever one exists.
5. Plan for change, so a new standard can be added without rebuilding content or reports.

This also answers the build versus existing components question. Build the parts that make your product different, and rely on existing standards and ready components for everything else. Product discovery services are a practical place to settle roles and data structure before you build, and an MVP development company will push you to ship the one standard your first customers need. A look at an online learning platform case study shows how scope decisions play out in a real product, and the wider question of how to develop an LMS without blocking future growth starts with the same role split.

For a scaleup the build versus reuse call comes down to differentiation. Build the learner experience, the data model and the reports that your customers pay for. Reuse ready components for the plumbing, such as a standard content player, a record store and an LTI library, because that work adds no visible value to your product. Your team keeps its time for the roadmap, and you avoid maintaining code that a standard already defines.

Ask four questions before any standard enters the roadmap. Which customers or partners require it in writing? Which data will you report on? Who owns the integration after launch? What happens to existing courses when the version changes? If you cannot answer all four, the standard is not ready for the roadmap.

Test with real customer packages and real tools before launch. Many LMS implementations interpret the specifications slightly differently, so a package that works in one system can fail in another.

Two mistakes are worth avoiding. Teams treat standards as interchangeable, and they add them too late. The fix for both is the same, which is to write down the role of each standard before the first sprint.