From 7f8e634a3d82f8f26ff139eb39a5bc3f9a5ca54a Mon Sep 17 00:00:00 2001 From: Chris Butler Date: Tue, 28 Apr 2026 17:46:56 -0700 Subject: [PATCH] FAQ: add enterprise billing, cost controls, and retry cost entries Adds three new FAQ entries under Costs & Usage covering: - Actions minutes vs AI usage billing distinction - Org-level spend controls and per-repo attribution - How retries and trigger frequency affect cost Source: github/agentic-workflows#495 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --- docs/src/content/docs/reference/faq.md | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/docs/src/content/docs/reference/faq.md b/docs/src/content/docs/reference/faq.md index 336b0fe8558..6c84c4417bc 100644 --- a/docs/src/content/docs/reference/faq.md +++ b/docs/src/content/docs/reference/faq.md @@ -546,6 +546,23 @@ Costs vary depending on workflow complexity, AI model, and execution time. GitHu Reduce costs by optimizing prompts, using smaller models, limiting tool calls, reducing run frequency, and caching results. +### Are GitHub Actions minutes billed separately from AI usage? + +Yes — these are two independent costs. Agentic workflows run as GitHub Actions workflows, so standard Actions minutes apply and are subject to your organization's [Actions spending limits](https://docs.github.com/en/billing/managing-billing-for-your-products/managing-billing-for-github-actions/managing-your-spending-limit-for-github-actions). AI usage is billed separately through your AI provider (see [Who pays for the use of AI?](#who-pays-for-the-use-of-ai) above). + +### How do I control and cap AI spend? + +gh-aw doesn't have a built-in spend cap. Cost governance happens at two levels: + +- **GitHub Actions minutes**: Set a spending limit in your GitHub organization's billing settings. +- **AI provider budgets**: For Claude and Codex, use Anthropic's or OpenAI's dashboard to set usage caps and alerts. For Copilot CLI, usage is constrained by the premium request quota in your Copilot plan. + +To attribute costs per repository, use a separate PAT or API key for each repo — then provider dashboards show usage scoped to that key. + +### How do retries and frequent triggers affect cost? + +Each workflow execution — including retries — runs a full agent session and consumes a similar amount of AI credits and Actions minutes. Agentic workflows don't loop indefinitely; the agent runs once per trigger. However, a high-frequency event trigger (e.g., firing on every comment across a busy repo) multiplies costs proportionally. Limit trigger frequency using conditions in the `on:` block of your workflow frontmatter, or restrict triggers to specific labels, paths, or actor filters. + ### Can I change the model being used, e.g., use a cheaper or more advanced one? Yes! You can configure the model in your workflow frontmatter: