Nothing is stored. No sign-up.
Next steps
One useful email every Monday
New free tools, AI and network threats that matter, and one quick challenge. No spam, unsubscribe in one click.
How to use it
- 1
Start with the biggest cost: input or output.
- 2
Turn on caching and batch where they fit.
- 3
Route easy tasks to a small model and re-check your bill.
Questions people ask
How can I reduce token usage?
Cache repeated prompt parts, send only relevant passages instead of whole documents, cap output length, trim chat history and send easy tasks to a smaller model.
How much does the Batch API save?
OpenAI, Anthropic and Google all charge 50% less for batch requests. OpenAI batches complete within 24 hours.
How much does prompt caching save?
Anthropic charges 0.1x the input price for cache reads on most models; OpenAI and Google list much lower cached-input prices too. A 5-minute Anthropic cache write costs 1.25x and pays off after one read.
Related tools
- Cheapest model for my taskPick a task and volume; see which model does it for the least per month.
- API bill estimator & budgetMonthly cost, caching and batch savings, and the day your budget runs out.
- AI subscription checkerWhich ChatGPT, Claude, Gemini or Copilot plans overlap, and what to cancel.
- Do I need a paid AI plan?Four questions tell you if Free, a $5–8 plan, $20 or $100+ fits.
- Free AI credits & student offersVerified student plans, startup credits, free tiers and hidden API credits.
Link to this tool
Found it useful? Paste this into your blog, wiki or runbook.
<a href="https://toolbini.com/ai/reduce-token-usage/">15 ways to cut token costs</a>Reviewed by the editorial team · Last reviewed 2026-11 · Report an error