n8n 2.42 Release Features: Prompt Caching Leads the List
A plain look at the n8n 2.42 release's features, led by prompt caching in the Anthropic node, following n8n's own 2.42.0 release notes.
Checked against the cited sources on .
Prompt Caching in n8n Workflows and What Else Changed in 2.42.0
If you've been hunting for a tidy rundown of the n8n release 26 features, the honest answer starts with one line in n8n's own 2.42.0 release notes: the Anthropic node gained prompt caching on its Message operation. That entry sits inside a long list of one-line titles n8n's GitHub release page groups under Features and Bug Fixes. The notes don't explain how to turn prompt caching on, what it costs, or how much latency it saves — the title is the whole entry.
That list-style format matters because it also covers changes to existing behavior, not just new options. Among the same day's Features entries, n8n's notes state that Agents are now enabled by default — a one-line change that can touch workflows a team already has running, long before anyone opens a changelog looking for AI news. Jan, n8n's CEO and founder, has described the company's approach to maintaining its many integrations this way, speaking generally on the Product Growth podcast:
We think skimming the full list of one-line titles is worth more than waiting for a feature to get its own headline, since a default behavior change can affect your workflows quietly, without a dedicated announcement. For teams already running Agent-type workflows, that one line is arguably more consequential than the caching option.
Sources: Release n8n@2.42.0 · n8n-io/n8n · GitHub
When n8n Announced the Update, and Which Version You're Actually Running
n8n's GitHub release page tags version 2.42.0 with the date 2026-09-29, which is the closest thing to an announcement date the official record gives for this update. That same page marks the release as a pre-release and does not separately state a general-availability date, so the tag date is what we're treating as when this change became official rather than a separate rollout date.
Searching for the n8n release 26 features won't turn up an exact tally in n8n's own documentation either. The archived 2.x release notes page we checked shows the current stable version as 2.42.6, a later patch than 2.42.0 — so your instance may already sit a few patches ahead, as of our check. That same archived page states, in its own words, that it is no longer updated, so it confirms version numbers without describing what any of them changed.
n8n also keeps a separate changelog it describes as a curated, narrative summary distinct from the full release notes and GitHub releases. The copy we found had no entry written specifically for version 2.42; the nearest dated narrative entries sit at 2026-09-30 and earlier versions such as 2.40 and 2.38. That gap is useful background, not a detail about 2.42.0 itself.
Sources: Release n8n@2.42.0 · n8n-io/n8n · GitHub, Release notes 2.x | Changelog | n8n Docs, Changelog | n8n Docs
Who n8n 2.42 Affects
Based strictly on what the 2.42.0 release notes state, three groups are affected. Anthropic node users can now turn on prompt caching for workflows that send Claude the same context repeatedly. Builders with existing Agent-type workflows need to know that Agents are now enabled by default, whether or not they asked for that change. And anyone running n8n at all is affected in the broader sense that this release also carries a long list of bug fixes alongside those two headline Features entries.
- Anthropic node users: anyone whose workflows call the Message operation can now use prompt caching.
- Agent-workflow builders: anyone with Agent-type workflows is affected by Agents being enabled by default.
- Anyone updating n8n: the release also bundles many unrelated bug fixes listed alongside the two headline features.
We think shipping caching to the Anthropic node first makes sense, since it's one of the integrations automation teams commonly reach for when they add an AI step to a workflow. That doesn't make the other two groups any less worth checking, though — a default change to Agent behavior is easy to miss if you're only watching for AI headlines.
The release notes we reviewed do not say whether these changes apply the same way on n8n Cloud as on a self-hosted instance, or whether community and enterprise editions differ here. The 2.42.0 release notes simply do not state this, so this article adds no claim about edition or plan differences from other sources.
Sources: Release n8n@2.42.0 · n8n-io/n8n · GitHub, Release notes 2.x | Changelog | n8n Docs
What to Do Now
None of this is instruction from n8n itself — the release notes list what changed without telling teams what to do about it. The suggestions below are our own, written for teams who want to move carefully rather than guess.
We'd treat prompt caching as an experiment rather than a swap-in upgrade for now: the release notes give it one line, with no stated setup steps, cost figures, or latency numbers, so a test run tells you more than the changelog does.
Sources: Release n8n@2.42.0 · n8n-io/n8n · GitHub, Release notes 2.x | Changelog | n8n Docs
What the n8n 2.42.0 Release Notes Don't Say
Staying strictly inside the official text, several things are simply absent. The notes don't explain how to enable prompt caching, what it costs, or how it affects response latency. They don't state an exact total for how many entries this release contains — so if you came here chasing a precise n8n release 26 features count, n8n's own notes don't give one, only a long list of one-line titles.
The notes also do not break any of this down by edition or hosted plan, a point already named above as not stated in the official text. The separate changelog we checked had no entry written specifically for 2.42, only neighboring dated entries — useful context, but not detail about this release.
Sources: Release n8n@2.42.0 · n8n-io/n8n · GitHub, Release notes 2.x | Changelog | n8n Docs, Changelog | n8n Docs
