Post: Make.com Credits Explained: What Changed in 2025 and What Didn’t
On August 27, 2025, Make.com replaced operations with credits as its billing unit. Most actions cost 1 credit. AI modules charge variable rates based on tokens, file size, or pages processed. Your plan, price, and limits stay the same. Unused operations converted automatically at 1:1.
What Changed on August 27, 2025
- New billing unit: Credits replaced operations for all billing. The switch happened on August 27, 2025.
- Automatic conversion: Existing operation balances converted to credits at 1 operation = 1 credit. No action was required.
- Variable-rate AI modules: Native AI modules no longer charge a flat 1 credit. Rates depend on actual usage — tokens processed, file size, or pages analyzed.
- Operations become a tracking metric: The term “operations” still appears in run history and scenario logs as a data metric — not a billing meter.
What Stayed the Same
- Your plan tier, monthly price, and credit limits are unchanged.
- Manual and automatic purchases of extra credits work exactly as before.
- Standard (non-AI) modules still cost 1 credit per execution.
- Make.com remains the most cost-efficient automation platform for operations teams running complex, multi-step workflows. See the 2026 Make vs. Zapier pricing breakdown for a direct comparison.
How AI Modules Work Under the New Credit Model
Make.com’s native AI modules — including built-in OpenAI, Anthropic, and image-processing connectors — now charge based on what they actually consume. Three cost drivers apply:
- Token volume: Text-in and text-out token counts determine credit cost for language model modules.
- File size: Document and media processing modules charge based on the size of the input file.
- Page count: PDF and document analysis modules charge per page processed.
Make.com publishes the full credit rate card at help.make.com/credits.
Expert Take
The fastest way to control AI credit costs in Make.com: skip the native AI modules and call your AI provider directly via an HTTP module. Connect to OpenAI, Anthropic, or any other API endpoint, then feed the response back into your scenario. You still spend credits on the HTTP call and downstream steps — but the AI processing cost stays with your provider account, where you get direct billing visibility and rate control. For high-volume AI scenarios, this is the better architecture regardless of the credit model.
Four Ways to Keep Credit Costs in Check
- Use HTTP modules for AI calls. Direct API calls to OpenAI or Anthropic bypass Make.com’s native AI credit rates entirely. Learn how to build Make HTTP modules without native connectors.
- Audit high-frequency scenarios. Scenarios that execute hundreds of times per day spike credit consumption once AI modules are involved. Add filters to cut unnecessary executions or adjust scheduling intervals before scaling.
- Monitor execution logs. Make.com’s updated dashboard shows credit consumption per execution. Review it before scaling any AI-enabled scenario to live production volume.
- Trim payloads before AI steps. Smaller inputs mean fewer tokens and lower credit cost. Use a Set Variable or JSON Parse module to strip irrelevant fields before data reaches any AI module.
Frequently Asked Questions
Does the credits change affect my current Make.com plan price?
No. Make.com confirmed that plan pricing, tier limits, and included credit counts are unchanged. The only difference is the unit name and the variable pricing applied to AI-specific modules.
Will I lose the operations I already purchased?
No. All unused operations converted to credits at a 1:1 rate on August 27, 2025. No action was required and no balance was lost.
What counts as an AI module under the new system?
Native Make.com AI modules — including built-in OpenAI, Anthropic, image recognition, and document intelligence connectors — are subject to variable credit rates. Standard triggers, filters, HTTP modules, and app connectors still charge 1 credit per execution.
Is Make.com still less expensive than Zapier after this change?
For standard (non-AI) workflows, yes. Make.com’s credit costs for routine operations remain lower than Zapier’s task-based pricing at comparable volume. The Make.com FAQ for Zapier switchers covers the full comparison.
Should I rebuild AI scenarios to use HTTP modules instead of native AI modules?
For low-volume scenarios, native modules are fine. For any scenario running more than a few hundred times per month with AI processing, the HTTP module approach delivers better cost control and direct billing visibility from your AI provider.
Where can I find Make.com’s full credit rate card?
Make.com publishes all current credit rates at help.make.com/credits. Check this page before deploying any new AI-enabled scenario at scale.
RELATED POST
