Send your content anywhere.
Build whatever you want on top.
One platform, two jobs. Dispatch SCORM packages to your customers' LMSs with seat caps, expiry, and revocation. Or build entirely new learning products on our hosted API — SCORM runtime, xAPI Learning Record Store, sub-tenant workspaces, all included.
# JSON in, hosted SCORM out curl -X POST https://api.allureconnect.com/v1/quick-publish \ -H "Authorization: Bearer ac_test_…" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "title": "Phishing Awareness · Q2", "spec": "scorm-2004-4", "lessons": [{ "title": "…", "questions": [...] }] }' ← 201 CREATED · 184ms { "id": "pkg_01HXFV9P", "status": "ready", "launch_url": "https://launch.allureconnect.com/pkg_…", "watermark": "sandbox" }
One API. The 12-month build you skip.
Native runtime support for every learning standard that matters. Your team ships in days, not quarters.
Send your content anywhere. Build whatever you want on top.
Whether your customers want SCORM-shell dispatched into their existing LMS, or you're building a new product on our learning primitives — Connect is the runtime underneath.
Send
Dispatch SCORM packages to your customers' LMSs — Cornerstone, Canvas, Moodle, WordPress, Workday — with seat caps, expiry, and revocation. Real-time completion data flows back to one dashboard.
- SCORM-shell dispatch to any LMS, your runtime stays in control
- Allowed-domain, seat-cap, and expiry guardrails on every route
- Per-client reporting · per-dispatch revocation · audit-grade evidence
- Embedded launch URLs for your own product or marketing pages
Build
Hosted SCORM runtime, xAPI Learning Record Store, and multi-tenant session management — exposed as clean REST endpoints. Sub-tenant API lets you run a white-label LMS on top of us.
- Quick Publish API — JSON in, hosted launch URL out
- Sub-tenant API for multi-tenant SaaS · parent/child workspaces
- MCP server & GPT Actions schema · listed on Anthropic registry
- Signed webhooks · idempotency keys · per-key spend caps
Train Acme on Cornerstone, Globex on Moodle, Hooli on Workday — from one source of truth.
Stop maintaining six versions of every course for six different LMSs. Publish once. Dispatch to your customers' systems with seat caps, expiry, and a kill switch. Watch completions flow back in real time.
One package. Many customers. Real licensing teeth.
Every dispatch route gets its own allowed domain, seat cap, expiry date, and revocation button. Your customer's LMS receives a tiny SCORM shell ZIP — we keep the runtime in your hands.
- →Allowed-domain enforcement · launches reject if referer doesn't match
- →Per-route seat caps and expiry dates; password challenges are planned
- →Republish a package, every active dispatch picks up the new version
- →One-click revocation — stops launches without touching the customer's LMS
Watch your content reach every learner — across every LMS — in real time.
One package → many destinations → completion callbacks return to one source of truth. Our dispatch fabric routes launches and reconciles xAPI records from every system your customers use.
The wedge: agent-native learning infrastructure.
We're the only learning platform with a published MCP server, GPT Actions schema, and sub-tenant API. If you're building with AI or shipping multi-tenant SaaS, you ship faster on Connect than on anyone else.
JSON in. Hosted SCORM out.
Send a structured course payload, get a launch URL back the same call. Idempotent. Versioned. Resumable. Watermarked in sandbox, clean in production.
- →Returns within ~200ms · validated, packaged, mounted
- →Re-publish to bump versions · old launch URLs keep working
- →Output works in any LMS that speaks SCORM
curl -X POST https://api.allureconnect.com/v1/quick-publish \ -H "Authorization: Bearer ac_test_…" \ -d '{ "title": "Phishing Awareness · Q2", "spec": "scorm-2004-4", "lessons": [{ "title": "Spotting suspicious senders", "questions": [{ "prompt": "Which domain is suspicious?", "choices": ["acme.com", "acme-secure.co"], "correct": 1 }] }] }' ← 201 Created { "id": "pkg_01HXFV9P", "launch_url": "https://launch.allureconnect.com/pkg_…" }
Stop duct-taping SCORM infrastructure together.
Most teams cobble together a half-dozen tools — SCORM Cloud, Rustici, a custom Postgres LRS, a homegrown dispatch shell, a webhook proxy. Connect replaces the whole stack.
What you're cobbling together today
- Three versions of every course — one for each customer's LMS quirks. Update once, you update everywhere. Manually.
- Email-chased completion CSVs from your customers' L&D admins every quarter, two weeks late, in inconsistent formats.
- No way to revoke access when a customer doesn't renew. The course just keeps running in their LMS forever.
- SCORM Cloud for hosted runtime — but $1k/yr minimums and per-launch fees that scale wrong for AI-builder traffic.
- Rustici Engine licensed and self-hosted on your infra — 6 months of integration before first learner.
- Custom Postgres LRS you built — works until xAPI verb cardinality blows up your queries.
- Webhook receivers, retry queues, dead-letter handling — three months of work, still flaky.
- One SCORM shell generator from a contractor — works for Cornerstone, breaks on Moodle.
- No MCP. No GPT Actions. No sub-tenant API. Every AI feature is yours to build.
What you actually need
- Hosted runtime from day one. Sandbox key in 30 seconds. Real production access on the $49 tier.
- Native dispatch for Cornerstone, Canvas, Moodle, WordPress, Workday, Blackboard. Shell generation handled.
- Built-in LRS with verb-aware indexing, anomaly detection, and audit-grade compliance evidence exports.
- Signed webhooks with automatic retry, dead-letter queue UI, and a replay button.
- MCP server, GPT Actions, sub-tenant API. The agent-native layer your competitors don't have yet.
- Per active learner. No tier cliff — overage gets cheaper as you grow. Dispatch, LRS & analytics bundled in one plan.
Pay per active learner.
One ladder, free to enterprise. A generous learner allotment on every plan, overage that gets cheaper as you grow, and no tier cliff. Dispatch, LRS, and analytics bundled in.
- 25 active learners / mo
- 5 GB storage
- Full API · MCP · GPT Actions
- No credit card
- 250 active learners / mo
- then $0.15 / extra learner
- Production keys, no watermark
- Email support
- 1,500 active learners / mo
- then $0.08 / extra learner
- Sub-tenant API · LTI 1.3
- Custom domain · 12h support
- 7,500 active learners / mo
- then $0.05 / extra learner
- Full white-label · prepaid credits
- 500 GB · 6h support
Trusted by content teams & developers
“We switched from SCORM Cloud and cut our content delivery costs by 60%. The API is cleaner and the dispatch system works exactly how we expected it to.”
“Our team had SCORM tracking running in production within a day. The developer sandbox and API docs made integration straightforward — no sales calls needed.”
“Finally, real-time completion data we can actually trust. The webhook events push straight into our data warehouse and the export-ready reports save us hours every month.”
Common questions
What is Allure Connect, in one sentence?
It's a hosted learning backend — SCORM runtime, xAPI Learning Record Store, dispatch engine, and API — that lets you ship content delivery and tracking infrastructure in days instead of months.
How is this different from SCORM Cloud?
SCORM Cloud is the incumbent; we're the modern alternative. We support every standard they do, plus newer ones (LTI 1.3, AICC HACP). We add an agent-native layer they don't have — MCP server, GPT Actions, sub-tenant API. We meter the same unit they do (active learners), but with no tier cliff — overage gets cheaper as you grow — and dispatch, an LRS, and analytics are bundled, not separate line items. And our sandbox + production keys are issued without a sales call.
What does "agent-native" mean and why should I care?
Two things: (1) our APIs are designed to be called by AI agents — idempotent, with structured JSON in, structured JSON out, watermarked sandboxes for safe agent traffic. (2) We're listed on the Anthropic MCP registry and ship a GPT Actions schema, so building "an AI that creates training courses" is a one-line config block, not a custom integration.
Can I run my own multi-tenant LMS on top of Connect?
Yes — the sub-tenant API unlocks on Growth and up. Your parent-scoped API key can mint and manage isolated child workspaces programmatically. Usage rolls up to one billing parent. Each child gets its own custom domain, branded player, and per-child quotas. Scale and Enterprise add full white-label.
Do I have to use your hosted runtime, or can I dispatch into my customer's LMS?
Both. Hosted launch URLs are good for embedding in your own product or AI workflows — Connect serves the runtime. Dispatch is for B2B scenarios where your customer hosts their own LMS and won't move — we generate a SCORM shell their LMS thinks is native, but launches call back to Connect. Most platforms use both.
What standards do you support?
SCORM 1.2, SCORM 2004 (all editions), xAPI 1.0.3 (Tin Can), cmi5, LTI 1.3, and AICC HACP. All six are first-class — no version detection, no spec coercion, no surprises.
How do you handle data residency?
Workspaces can be pinned to us-east, us-west, eu-west, or ap-southeast. Statements stay in the region you pick. Encryption at rest and in transit by default. SOC 2 in progress.
What happens if my agent goes haywire and burns through my quota?
On paid plans we never block your learners — caps are soft, so a traffic spike just bills overage and your product keeps working. We alert you at 70%, 90%, and 100% of your allotment, and Scale/Enterprise can set a prepaid-credit ceiling that hard-stops with no bill shock. Per-key burst limits (default 500 req/s, sliding window) still protect against runaway loops.
Start sending or start building today.
Free developer sandbox. Full API access from day one. Production keys on the $49 tier.