# Pay-per-use API tools vs subscriptions for AI agents Compare pay-per-use API tools with subscriptions by workload, billing unit, provider-account overhead and spending predictability. By [Scrollport](https://scrollport.com/) Published 3 August 2026 Updated 10 September 2026 In brief ## When is pay-per-use better than a subscription for an AI agent's API tools? Pay-per-use API tools fit irregular agent work because the bill follows searches, results, verifications or generated assets rather than a fixed seat. Subscriptions can be better for stable, repeated use of one provider. Scrollport exposes each catalog tool's billing unit and current price before execution, then settles successful work against one prepaid wallet. Pay-per-use API tools fit irregular agent work because the bill follows searches, results, verifications or generated assets. A subscription can be better when the same provider is used continuously and the included allowance is predictable. The choice depends on workload, billing unit, account overhead and the controls available before execution. ## Pay-per-use API tools vs subscriptions Per-seat subscriptions are useful when a team is buying a stable workspace for humans or a predictable allowance from one provider. They are a weaker fit for an agent that may run once today, run a hundred times tomorrow and sit idle next week. The subscription remains fixed while the work varies, which can make occasional cross-provider jobs expensive and harder to attribute. A usage-based model starts with a clearer question: what unit did the work consume? That unit might be a search, result, verification, generated image or another provider-defined event. The denominator matters. A price shown as “per call” when the provider bills per result would make a cheap-looking call impossible to reason about. ## Usage pricing maps the bill to the job Scrollport keeps the public taxonomy explicit. A **category** is a broad navigation group. A **capability** is a provider-neutral job or outcome. A**provider** identifies the recognised maker, lab, connected app or marketplace, and a **catalog tool**is the independently discoverable, inspectable, priceable and runnable operation. The current [web research and extraction tools](https://scrollport.com/tools/categories/web-research-extraction) show the kind of job-led surface an agent can browse without choosing a supplier first. This makes a bill legible: the agent selects a catalog tool for a capability, the tool exposes its billing unit, and the wallet settles the measured usage. A reader can compare work rather than comparing opaque plan tiers. The live catalog owns current tool descriptions and prices, so this article deliberately does not repeat a provider rate that could change. ## What is a one-wallet API catalog? A one-wallet API catalog lets an agent discover and pay for different specialist APIs through one funded account rather than maintaining a balance with every provider. The catalog must still expose each tool’s real billing unit and current price; one wallet should simplify payment, not make unlike units look identical. One Scrollport credential gives an agent one authentication path into the nine control tools:`list_apps`, `search_tools`, `inspect_tool`, `run_tool`, `get_run` and `get_wallet`. One wallet gives the human one place to top up, set a spending boundary and review what was charged. That is simpler than teaching an agent several provider billing portals and several account-level limits. Scrollport’s core tools include provider access, so they need no new provider account, API key or subscription. A connected-app tool is a separate exception for acting inside a service the human already uses. Authorise that app once through Scrollport so the agent does not need its API key or OAuth token in a prompt, configuration or log. ## The boundaries of “pay for what you use” Usage-based pricing is transparent only when its boundaries are visible. The catalog should state the billing unit and current published price, and a run should reserve the expected amount before execution. A successful run captures the actual cost and releases any remainder; a failed run releases the hold rather than charging for work that did not settle. - It is not a promise that every provider connection is free or unnecessary. - It is not a promise that a provider’s public rate never changes. - It is not a guarantee that every job has the same unit or the same cost. - It is not permission to hide failed calls, holds or released funds. The practical rule is simple: describe the work, expose the denominator, keep the live catalog authoritative and let the wallet show the money movement. For the full setup and control-tool path, [read the platform documentation](https://scrollport.com/docs); for the product model, read [why Scrollport uses a small control surface](https://scrollport.com/resources/why-scrollport-uses-a-small-control-surface). ## Related reading [Agent spend controls AI agent API spending controls: approval thresholds and daily caps Follow a Claude Code-compatible walkthrough of tool prices, a settled search charge, human approvals, daily caps and optional auto top-up.](https://scrollport.com/resources/control-ai-agent-spending) [Agent tool access How to give AI agents API access without provider accounts or keys See a sourced company-research example with actual tool costs, then compare API access models and optional connected apps.](https://scrollport.com/resources/give-ai-agents-access-to-external-tools) --- Canonical page: https://scrollport.com/resources/pay-per-call-ai-agent-tools