n8n node library: the 22 most used nodes, with data
n8n's docs tell you what each setting does. This library adds what the docs don't: how often 13,574 public templates use each node, which nodes it is paired with, how it has changed on GitHub, and the problems people hit, each with a source.
All 22 nodes
Click a column heading to sort. "Usage" is the share of the 13,574 templates that contain the node. "AI share" is the share of those templates that also contain an AI node; for AI nodes it is 100% by definition. "Strong partners" counts nodes that turn up next to it far more often than average.
| # | Node | Category | Score /80 | Usage | Templates | AI share | Strong partners | Trouble signal |
|---|---|---|---|---|---|---|---|---|
| 1 | Simple Memory | AI | 71.7 | 8.0% | 1,086 | 100.0% | 22 | no trouble signal |
| 2 | AI Agent | AI | 69.9 | 31.3% | 4,246 | 100.0% | 13 | no trouble signal |
| 3 | OpenAI Chat Model | AI | 69.6 | 22.9% | 3,104 | 100.0% | 13 | no trouble signal |
| 4 | Structured Output Parser | AI | 68.3 | 16.1% | 2,181 | 100.0% | 14 | no trouble signal |
| 5 | Chat Trigger | Triggers | 68.1 | 5.6% | 761 | 100.0% | 19 | no trouble signal |
| 6 | Google Gemini Chat Model | AI | 65.1 | 9.5% | 1,285 | 100.0% | 14 | no trouble signal |
| 7 | Switch | Logic | 59.3 | 15.2% | 2,069 | 65.7% | 12 | no trouble signal |
| 8 | Split Out | Data | 58.3 | 11.6% | 1,577 | 60.9% | 13 | no trouble signal |
| 9 | Merge | Logic | 55.1 | 19.3% | 2,616 | 64.3% | 6 | no trouble signal |
| 10 | Telegram | Apps | 54.9 | 12.5% | 1,691 | 69.5% | 7 | no trouble signal |
| 11 | Loop Over Items | Logic | 54.0 | 17.1% | 2,319 | 57.0% | 7 | no trouble signal |
| 12 | Gmail | Apps | 53.3 | 22.5% | 3,059 | 65.8% | 3 | no trouble signal |
| 13 | Schedule Trigger | Triggers | 52.8 | 31.8% | 4,315 | 53.5% | 4 | no trouble signal |
| 14 | Google Sheets | Apps | 52.2 | 35.7% | 4,846 | 58.3% | 2 | no trouble signal |
| 15 | Slack | Apps | 51.6 | 16.6% | 2,252 | 60.4% | 4 | no trouble signal |
| 16 | Edit Fields (Set) | Data | 51.3 | 55.7% | 7,554 | 58.1% | 0 | no trouble signal |
| 17 | Wait | Logic | 51.2 | 17.7% | 2,398 | 55.5% | 4 | no trouble signal |
| 18 | Code | Data | 51.0 | 59.4% | 8,061 | 55.0% | 0 | no trouble signal |
| 19 | Webhook | Triggers | 51.0 | 19.7% | 2,676 | 51.0% | 4 | no trouble signal |
| 20 | HTTP Request | Data | 50.5 | 51.6% | 7,002 | 51.0% | 1 | no trouble signal |
| 21 | n8n Form Trigger | Triggers | 50.5 | 11.5% | 1,559 | 59.7% | 5 | no trouble signal |
| 22 | If | Logic | 49.6 | 50.5% | 6,849 | 53.0% | 0 | no trouble signal |
Telegram combines the Telegram node and the Telegram Trigger, which often work as a pair; the page shows each count separately.
Top 5 by score
- Simple Memory: 71.7 of 80 (8.0% of templates, 100.0% AI share, 22 strong partners)
- AI Agent: 69.9 of 80 (31.3% of templates, 100.0% AI share, 13 strong partners)
- OpenAI Chat Model: 69.6 of 80 (22.9% of templates, 100.0% AI share, 13 strong partners)
- Structured Output Parser: 68.3 of 80 (16.1% of templates, 100.0% AI share, 14 strong partners)
- Chat Trigger: 68.1 of 80 (5.6% of templates, 100.0% AI share, 19 strong partners)
How the score works
Score = 40 × U + 20 × A + 20 × P, out of 80. U is the node's usage rank: the share of the 126 node types that appear in at least 50 public templates and are used in no more templates than this one (1.0 = the most used). A is the share of the node's templates that also contain an AI node. P is how many partner nodes it is strongly paired with (in at least 5% of its templates and at least 1.5 times more often than across the whole library), divided by the highest count among these 22 nodes (22). The planned fourth part, T (how often the node shows up in forum trouble threads), is not measured, so it is left out rather than guessed.
- Measured: U, A and P, all from our own count of public templates. We store totals only.
- Not measured: T, the trouble signal. We planned to count forum threads per node, but the n8n community forum's terms of service forbid automated access and monitoring, so we didn't collect forum data. Every node shows "no trouble signal" for that reason. It does not mean the node has no problems; see each page's known problems.
- Not used: npm download counts (n8n ships its nodes in two big packages, so downloads can't be split per node), GitHub issue counts (shown only as context), and the template counts on n8n.io integration pages.
- What it means: a high score says a node is used widely and sits at the centre of many builds, especially AI ones. It is not a quality rating. Public templates show what people publish, not what runs in production.
Why 22: they are the most used nodes in the snapshot, after leaving out the sticky note, the manual trigger and No Operation, which don't do work. Each page needed its own usage data, its own sourced problems and its own GitHub history; all 22 had enough, so none was folded into this page.
Related
- 25 free n8n templatesImport-ready workflows that use these nodes
- The n8n template studyThe full 13,574-template dataset these numbers come from
- n8n vs Zapier vs MakeWhat the same agent costs on each
- n8n community nodesInstall paths, risks and the n8n 3.0 change
- Most-used community nodesRanked monthly by downloads and retention
- n8n use casesWhat public templates are built to do, by category
Questions
- What are the most used nodes in n8n?
- In public templates: Code (59.4%), Edit Fields/Set (55.7%), HTTP Request (51.6%) and If (50.5%). Google Sheets is the most used app node.
- How is this different from the n8n docs?
- The docs explain settings. These pages add counts from 13,574 public templates, pairings, version and GitHub change history, and known problems from the docs and merged fixes, each linked. Always check the docs for how a setting works.
- Why does every node say no trouble signal?
- Because we didn't measure it. The n8n forum's terms forbid automated access, so we left the forum part out of the score rather than estimate it.
- Is a higher score better?
- No. It measures how central a node is in published templates, not how good or reliable it is.
- How often is the data updated?
- The template snapshot is from 2026-10-20; versions, docs and GitHub history were pulled Oct 26, 2026. Each page shows its dates.