Use a small model for simple jobs and save the big one for hard ones
Most bot jobs (sorting email, summarizing, checking a page) don't need the most powerful model. The small models cost a fraction as much, so route easy work to them and only send hard problems up.
What the bot does
- Sends routine work (summaries, sorting, yes/no checks) to a small, cheap model.
- Sends hard work (tricky reasoning, long documents, important drafts) to a bigger model.
- Optionally lets the small model decide: if it isn't sure, it hands the job up.
- Logs which model handled what, so you can see where the money goes.
What you end up with
The same results on everyday jobs for a fraction of the cost, with the expensive model kept for the jobs that need it.
Copy-paste prompt
Fill in the [brackets]. Then paste it into any tool below.
You are a triage step. Read the task below and answer with one word: SIMPLE (summarize, sort, extract, yes/no) or HARD (multi-step reasoning, legal/medical/financial wording, long documents, anything you're unsure about). Task: [paste task].
Tools that can do it today
- automationn8nRoute steps to different models in one workflowFree to self-host; paid cloud plans
- automationMakeFree plan; paid plans add more
- automationZapierFree plan; paid plans add more
- agentGrok Bot (personal agents)Pick the model per bot where the app allows itCheck plan requirements in xAI's docs
Plans and limits change; check each tool's page.
Watch out
- Test the small model on 10 real examples before switching. Cheap is no good if it gets things wrong.
- Prices change. Anthropic's small Haiku 5.5 lists $0.10 / $0.50 per million tokens and its top Fable 5.1 $10 / $50 (checked Oct 21, 2026); see our price table for the latest.
Questions people ask
Which AI tools can do this?
n8n, Make, Zapier or Grok Bot (personal agents) can do this today. Paste the copy-paste prompt on this page into any of them and fill in the [brackets].
How hard is it to set up?
It's rated Medium. Start with the steps under “What the bot does”.
What should I watch out for?
Test the small model on 10 real examples before switching. Cheap is no good if it gets things wrong.
Sources
- docsAnthropic pricing docs: "Choose Haiku for simple tasks, Sonnet for most production workloads, and Opus for the most complex reasoning"
- docsOpenAI API pricing (official)
- docsGemini API pricing (official)
Last checked Oct 21, 2026. Self-reported results from posts are the poster's own claims.