LearningFlow LearningFlow vs n8n
Both products lay work out as nodes on a canvas, and for connecting business systems n8n has the deeper catalogue of ready-made nodes and the longer track record. LearningFlow is built for the other half of the job, where the flow is also the lesson a class works through and every flow has an address your own software can call.
Automation · Sources last read 13 September 2026
Choose n8n if
Choose n8n when the job is plumbing between systems of record, and nobody outside IT ever opens it.
- Syncing a student information system with the LMS, finance and a CRM every night.
- Wiring ten business systems together where a ready-made node already exists for eight of them.
- An ops team that wants workflows held in git and promoted from a development instance to production.
Choose LearningFlow if
Choose LearningFlow when the same canvas has to produce a tool staff run and an assignment a class completes.
- A rubric flow a head of department runs, then sets as a Year 7 assignment.
- A teacher leaves in July and the school keeps the tools she built.
- You want a wrong connection refused as you drag it, before anything runs.
Where the two are built differently.
Nodes that run behind the scenes.
The canvas is excellent and its audience is machines. Nothing here is something you hand to a person and ask them to work through.
The same canvas. Two audiences.
Connections are typed, so the canvas refuses a wrong wire while you are dragging it rather than at three in the morning. And every flow has an address, so your own software can run it.
What each one does.
Written from each product's own documentation. Where something is true only on a higher tier, the row says so.
| n8n | LearningFlow | |
|---|---|---|
| Who logs in | The account types are owner, admin and member, with no learner account, no join code and no assignment. | Students are a first-class user type with their own accounts, join codes and assignments. |
| Switching features per person | Roles are owner, admin and member across the instance, plus admin, editor and viewer inside a shared project, and no feature switches on or off for one person. | Seven capabilities switch per member: org-library, chat, flow-creator, assignment-creator, solution-designer, base and work-with-ai. |
| If the builder leaves | A workflow in a personal project belongs to its creator, and an admin either moves it into a shared project or transfers it when that user is deleted. | The organisation id is stamped at creation and never moves, so removing a member deletes the membership row, not the work. |
| Breadth of connectors | Hundreds of built-in app nodes, plus an HTTP Request node and a Code node for anything without one. | Published apps call a named list: Gmail, Resend, WhatsApp, Firecrawl, document parse, web search and crawl, speech to text, image and video generation. |
| Development to production | Workflows can be held in git and promoted from a development instance to production through environments. | There is no git-backed promotion from a development instance to a production one. |
| What comes out as files | The Convert to File node writes CSV, XLS, XLSX, ODS, JSON, HTML, RTF, text and ICS, so a PowerPoint or a Word file means calling an outside service. | Work with AI returns .pptx, .docx, .xlsx, .mp4, .mp3 and .png, and a spreadsheet comes back with working formulas. |
| Calling it from your code | A public REST API covers workflows, executions, credentials, users and projects, authenticated with an X-N8N-API-KEY header. | POST /task/:taskId with an x-api-key, optional SSE streaming, and per-key credit top-up and usage reporting. |
| Learning records | There is no learning record concept, though an HTTP node could post statements to a store you already run. | xAPI package ingest, a self-hosted SQL LRS, or point at the organisation's own LRS. |
What comes with it here.
A real Linux machine
Work with AI runs Python 3.12, pandas, matplotlib and LibreOffice in a sandbox with the network switched off.
A narrated whiteboard video
One sentence becomes an MP4 in 16:9, 9:16 or 1:1, with subtitles off, burned in, or as a separate .srt file.
Per-app database contracts
Several apps share one base, and each holds its own contract over named tables, named columns and a row filter, with an expiry.
59 ready-made teacher tools
Lesson plans, marking rubrics, quizzes, worksheets, report comments, text levelling and YouTube questions sit in a public gallery.
Where n8n is the better choice
If the real requirement is integration plumbing, buy n8n. An IT team syncing a student information system with the LMS, the finance system and a CRM will find a ready-made node for most of it, and an HTTP Request node or a Code node for the rest. LearningFlow has nothing like that catalogue or that community. The self-hosted Community Edition is also free, so a technical team can have that node catalogue running on its own server this afternoon. The paid tiers add SSO, shared projects and the git environments.
The whole stack against n8n.
The table above compares the two at the job they share. This one runs every capability past both, because what a school is choosing between is usually a tool that does one job and a workspace that does many. Hover any mark to see what it means, and a partial one to see exactly what is limited.
| LearningFlow | n8n | |
|---|---|---|
| Ask and read | ||
| Chat with AI | ||
| Research the web and write it up | ||
| Read text off an image or a scan | ||
| Agent browses the live web | ||
| Make | ||
| PowerPoint deck | ||
| Spreadsheet with working formulas | ||
| Word document | ||
| Image generation | ||
| Video generation | ||
| Narrated whiteboard video, corrected scene by scene | ||
| Voice-over | ||
| Song generation | ||
| Interactive web page | ||
| Edit what it made, online | ||
| Slide and design templates | ||
| Presenter on camera, from a photo you supply | ||
| Teach | ||
| Ready-made teacher tools | ||
| Build your own tool, no code | ||
| Lesson that branches on the answer | ||
| Hears them read aloud, checks pronunciation | ||
| Assign to a class with a join code | ||
| Assignments and submissions | ||
| Student accounts | ||
| LTI link from your LMS | ||
| Learning records to your own LRS | ||
| SCORM 1.2 package out | ||
| Build | ||
| Working app | ||
| Full working platform templates | ||
| Several apps over one database, rights each | ||
| App sends email or WhatsApp | ||
| App runs on a schedule | ||
| App accepts webhooks | ||
| Generate real-time apps | ||
| Generate AI-driven apps | ||
| Shared database | ||
| Revoke one app's access, not the others | ||
| Run | ||
| Run any tool from your own software | ||
| Choose the AI model, on every plan | ||
| Switch a feature off per group | ||
| Organisation keeps the work when staff leave | ||
| We do not do this | ||
| Transcribe a video or audio file | ||
| Custom domain for a published app | ||
Before they decide.
Can we run n8n and LearningFlow together?
Yes, and most buyers should. Keep the system-to-system syncs in n8n, and have an HTTP Request node post to /task/:taskId with an x-api-key to run a LearningFlow flow. The response can stream, and each key carries its own credit top-up and usage report.
Can we import our existing n8n workflows?
No. There is no importer, and the two canvases do not share a format. Rebuild the flows a teacher or a student actually touches, and leave the nightly data syncs where they already work.
What will each one cost us?
n8n Cloud Starter is 20 euros a month billed annually for 2,500 executions with unlimited users, Pro is 50 euros a month annually for 10,000, and self-hosted Business is 667 euros a month annually for 40,000. Every plan counts executions rather than users, so the bill follows how often the flows run. LearningFlow charges credits with no subscription required, and personal, organisation and enterprise wallets stay separate, so a teacher's free monthly grant never pays for the school's work.
If the person who built everything leaves, do we keep it?
In n8n, a workflow built in a personal project belongs to its creator until somebody moves it into a shared project. Deleting that user prompts you to transfer their workflows and credentials to another person. Shared Projects avoid this but are capped at one on Starter, three on Pro and six on Business. In LearningFlow the organisation id is stamped at creation and never moves, and an admin can reassign org-library tools whose owner has left.
Will either one hand us a finished deck or document?
n8n writes CSV, XLS, XLSX, ODS, JSON, HTML, RTF, text and ICS itself, and reaches an outside service such as OpenAI for an image or an audio file. A deck or a Word report means calling an API you arrange. LearningFlow returns .pptx, .docx, .xlsx with working formulas, .mp4, .mp3 and .png. Its sandbox has no network, and it cannot read a scanned page or watch a video.
Where this came from. Read on 13 September 2026. n8n changes its product and its prices without telling us, so check anything that decides a purchase.
Try it on your own work.
Teachers get credits every month, free. No card, no seat count.