Claude Code Routines is included in your existing Claude Pro ($20/mo), Max ($100-$200/mo), or Team plan -- no extra line item. The constraint is per-day run caps: 5 runs on Pro, 15 on Max, 25 on Team. For AI-reasoning-heavy workflows that stay under those caps, Routines can eliminate your separate automation tool subscription entirely.
I've been running Routines since they shipped in April 2026. The economics surprised me -- not because Routines is obviously cheaper, but because the comparison isn't straightforward. Here's how I actually think about when to use it vs. keeping n8n.
What are you actually paying for with Claude Code Routines?
Claude Code Routines runs on Anthropic's infrastructure with no additional subscription cost beyond your Claude plan. Pro at $20/month gets 5 routine runs per day. Max ($100 or $200/month) gets 15. Team or Enterprise seats get 25. Each routine run draws down your existing subscription's token budget the same way an interactive session would -- no separate billing line, no new dashboard.
Routines launched in April 2026 in research preview, available across all paid plans: Pro, Max 5x, Max 20x, Team, and Enterprise. You create them at claude.ai/code/routines or with /schedule in the CLI. Your Routines usage rolls into the same 5-hour token window as everything else.
The daily run cap is the economic constraint that matters most. If you need more than 5 recurring automations per day on Pro, you either hit the cap and wait for reset, or upgrade. For most solo developers running background automation -- nightly code digests, daily PR reviews, weekly reports -- 5 runs on Pro is enough. For production environments where multiple Routines fire across a team, the 25-run Team cap is the real ceiling.
How does the daily run cap actually work?
The daily cap applies specifically to scheduled recurring routines, not one-off runs. A one-off run -- triggered via API call or manually -- doesn't touch the daily routine cap; it draws from your standard subscription usage like any interactive Claude Code session. The cap tracks how many recurring automated routine executions fire per account per day.
A practical example: On Pro, you get 5 routine slots per day. A morning PR digest, a nightly test runner summary, a daily changelog scan, a weekly metrics report (fires once on Friday, not daily), and a Tuesday competitor alert. That's 5 recurring routines -- exactly at the Pro cap. Add a sixth and you need to upgrade or remove something.
The Max cap of 15 is enough for most serious solo developers. The 25-run Team and Enterprise cap covers most small-team production deployments. One nuance worth knowing: Team and Enterprise accounts with extra usage enabled can run Routines past the daily cap on metered overage billing. Pro and Max users who hit the cap have additional runs rejected until the window resets -- there's no metered safety net on individual plans.
How Claude Code Routines stacks up against n8n on price
n8n Cloud pricing in 2026 runs $20/month for Starter (2,500 workflow executions/month), $50/month for Pro (10,000 executions), and around $667/month for Business (40,000 executions). Self-hosted n8n is free software -- your real cost is the VPS, typically $10-20/month for a reliable instance. n8n charges per workflow execution: a 10-step workflow that runs 1,000 times costs 1,000 executions, same as a 1-step workflow running 1,000 times.
The comparison isn't direct because n8n and Routines solve different parts of the problem. n8n is an orchestration layer: it connects APIs, moves data, and conditionally branches logic. Claude API calls in n8n are a separate line item -- you pay n8n Cloud for the orchestration plus Anthropic API tokens for the intelligence. Routines bundles both into your existing subscription.
A developer who replaced their n8n Cloud Pro workflow stack with Routines in 2026 estimated annual savings of $660-$780 just on the subscription cost -- dropping the $50/month n8n Pro line while keeping the same Claude Max subscription. The math holds when your recurring automation count stays below 15/day and Claude is already doing the reasoning work in those workflows. If you're paying $200/month for Max plus $50/month for n8n to coordinate Claude API calls, Routines cuts that second line to zero.
Want this built for your business?
Venti Scale builds AI automation systems for businesses that want results without the learning curve. One operator, AI-powered, full marketing stack.
See What We Build
When n8n still wins on price
n8n's execution-based billing wins at volume. If you need 50+ automation runs per day -- high-frequency webhook processing, real-time event handling, continuous integration pipelines -- Routines' daily cap blocks you. The maximum on Team/Enterprise is 25 scheduled routine runs per day. n8n Cloud Starter's 2,500 executions/month works out to roughly 80 per day, and self-hosted has no cap at all.
n8n also wins when Claude isn't the intelligence layer. Moving data between APIs, transforming JSON, syncing CRM records, processing webhooks without AI reasoning -- those workflows have no business using Claude tokens. n8n's visual node builder handles them faster to build and cheaper to run at scale. No token cost, no daily cap, just execution counts.
The self-hosted n8n argument is particularly strong for high-volume, low-complexity automation. A $15/month VPS running n8n handles thousands of executions per day with zero per-run cost. For that use case, Routines isn't even in the conversation. Where Routines gets interesting is specifically in the "I'm already paying for Claude and I need fewer than 15 smart automations per day" scenario.
The hidden cost both tools skip: build time and maintenance
Subscription price is one number. Build time and ongoing maintenance are the numbers most cost comparisons ignore. A complex n8n workflow that coordinates Claude API calls -- PR review, comment generation, Slack notification -- typically involves 4-6 nodes, a custom function node with the prompt, webhook setup, and a Claude API credential. That's a 2-3 hour build plus ongoing maintenance when the n8n version upgrades or the Claude API schema changes.
The equivalent Routines implementation is a Claude Code prompt with a trigger. The same PR review routine that took 3 hours in n8n takes 20-30 minutes in Routines. If a developer charges $100/hour for their time, the n8n build costs $300. The Routines build costs $50. That's a $250 hidden cost advantage on day one -- before any subscription math.
The GitHub trigger support in Routines is particularly relevant here. Supported GitHub events include PR creation, pushes, issues, check runs, workflow runs, discussions, releases, and merge queue events. Wiring any of those in GitHub Actions requires a YAML workflow file, secrets configuration, a job runner, and a separate Claude API step. In Routines, it's a prompt plus a trigger selection. The connectors -- Slack, Linear, Google Drive, GitHub -- handle authentication automatically once connected.
How to calculate your actual break-even point
The break-even calculation for switching from n8n to Routines comes down to three questions. First: how many distinct recurring automation tasks do you run per day? If the answer is above 15, you need Team seats ($100/seat, 5-seat minimum, $500/month floor) or you need n8n. At that price floor, n8n Cloud Pro at $50/month is dramatically cheaper for pure workflow volume.
Second: does Claude currently appear in those workflows? If your automations already make Claude API calls, you're paying Anthropic twice -- once for your Claude subscription and once for API tokens in n8n workflows. Routines collapses that into a single cost. If your automations don't use Claude, Routines adds zero value and you should stay on n8n.
Third: what plan are you on? If you're on Claude Pro ($20/month) and you're also paying n8n Starter ($20/month) for 5 or fewer automation tasks, Routines eliminates the n8n line -- saving $240/year with no capability loss. If you're on Max ($200/month) and running n8n Pro ($50/month), the savings are $600/year assuming your workflow count fits the 15-run cap. If those savings matter and your volume fits, the switch is straightforward. If you need more than 15 runs per day or your automations are Claude-free, n8n wins.
FAQ
Is Claude Code Routines free to use?
Routines is included in Claude paid plans -- Pro ($20/month), Max ($100-$200/month), Team, and Enterprise -- at no additional charge. The constraint is a daily run cap: 5 runs/day on Pro, 15 on Max, and 25 on Team/Enterprise. Runs past the cap are rejected on individual plans or billed as metered overage on Team/Enterprise accounts with extra usage enabled.
Can Claude Code Routines replace n8n?
Routines can replace n8n for AI-reasoning-heavy workflows running fewer than 15 times per day that already use Claude as the intelligence layer. For high-volume, high-frequency automations, data-movement workflows without AI reasoning, or any use case above the Routines daily cap, n8n's self-hosted free tier or Cloud Starter at $20/month remains more cost-efficient.
What triggers does Claude Code Routines support?
Routines support three trigger types: scheduled (cron-based time intervals), GitHub events (PR, push, issue, check run, workflow run, discussion, release, and merge queue events), and API/webhook (HTTP POST to a dedicated endpoint with bearer token). Connected integrations include Slack, Linear, Google Drive, and GitHub for reading context and posting outputs.
What happens when you hit the Routines daily run cap?
On Pro and Max plans, routine runs that exceed the daily cap are rejected until the next reset period. One-off API-triggered runs are exempt from the daily cap -- they draw from standard subscription usage. Team and Enterprise accounts with extra usage enabled can continue running past the cap on metered overage billing. Solo plan users have no metered fallback and must wait for the cap to reset.
Does Claude Code Routines work on the free Claude plan?
No. Routines requires a paid Claude plan: Pro, Max, Team, or Enterprise. The free Claude plan does not include Claude Code access, and Routines is a Claude Code feature. It launched in April 2026 in research preview. Access starts at $20/month on Pro with 5 routine runs per day.
Want this built for your business?
Venti Scale builds AI automation systems for businesses that want results without the learning curve. One operator, AI-powered, full marketing stack.
See What We Build
