GitHub Copilot and Claude Code Promotions End This Week: A 30-Minute Cost Check

Two AI coding allowances step down within the same week. GitHub says the higher promotional pool for existing Business and Enterprise customers covers June, July, and August; at the 1 September 2026 monthly reset at 00:00:00 UTC, included usage returns to the standard amounts. Claude Code’s temporary 50% increase to eligible weekly limits ends on 31 August 2026 at 11:59 PM PT. For a freelancer or small team using both, the immediate job is to check the standard allowance and the overage switch before the next heavy coding session.

Plan, allowance, and overage details below were checked on 28 August 2026. Confirm them again on each vendor’s official billing page before changing a subscription.

Check overage first: GitHub enables additional usage by default for organizations and enterprises; Claude’s individual paid-plan credit flow begins with an explicit Enable step and a prepaid balance.

What changes at each deadline

Service	Allowance after the promotion	What stays the same	Overage posture to check
GitHub Copilot Business	1,900 AI credits per assigned seat each month, pooled at the billing-entity level	Monthly seat pricing; unlimited code completions and Next Edit suggestions on paid plans	Additional usage is enabled by default for organizations and enterprises
GitHub Copilot Enterprise	3,900 AI credits per assigned seat each month, pooled at the billing-entity level	Monthly seat pricing; unlimited code completions and Next Edit suggestions on paid plans	Additional usage is enabled by default for organizations and enterprises
Claude Code on eligible plans	Standard weekly limit after the temporary 50% increase ends	The rolling 5-hour limit, plan, and billing	Individual Pro and Max usage credits require enablement and prepaid funds; Team and Enterprise users should verify organization controls in their live account

GitHub’s organization and enterprise billing documentation, accessed 28 August 2026, lists the standard amounts above, the promotional amounts of 3,000 for Business and 7,000 for Enterprise, and a promotional period of 1 June–1 September 2026. It also says the monthly pool resets at 00:00:00 UTC on the first day of each calendar month. GitHub’s transition announcement, also accessed 28 August 2026, defines the promotional months as June, July, and August and says Business and Enterprise monthly seat pricing stays unchanged.

Anthropic’s promotion terms, accessed 28 August 2026, apply the 50% weekly increase to Claude Code for Pro, Max, Team, and legacy seat-based Enterprise users. The temporary increase covers Claude Code across the CLI, IDE extensions, desktop, and web. The regular 5-hour usage limit keeps its existing cadence, and the weekly limit returns to standard after the promotion. Plan and billing remain unchanged.

The cost-control difference that matters

One GitHub AI credit equals $0.01 USD. Chat, CLI, cloud agents, Spaces, Spark, and third-party coding agents can consume credits according to token use and model pricing. Code completions and Next Edit suggestions consume zero AI credits and remain unlimited on paid plans. This split gives a small team a practical fallback inside the same seat: keep inline help available while governing the metered agent and chat work.

GitHub pools included credits at the billing-entity level. A heavy user can draw more than the per-seat amount while lighter users offset that usage. The same pool can also hide an individual spike until the pooled balance is gone. GitHub then follows the organization’s additional-usage policy, which starts enabled by default. An administrator who wants a hard stop must disable the AI credits paid-usage policy in AI Controls. Teams that allow overage can set user, cost-center, organization, and enterprise controls according to their account structure.

Claude uses an explicit activation path for individual Pro and Max usage credits. Anthropic’s usage-credit guide, accessed 28 August 2026, directs an individual paid-plan user to open Settings → Usage, click Enable, add a payment method, configure spending preferences, choose a monthly cap or unlimited spending, and prepay funds. Auto-reload is optional. Once the plan limit is reached and credits are available, the user can choose to continue; subsequent usage is billed at standard API rates and the regular 5-hour reset timing continues. Check the current Claude pricing page for the plan and usage rates shown to your account before enabling paid usage.

That difference changes the first question in an audit. On Copilot Business or Enterprise, ask, “Where is the hard stop?” On an individual Claude paid plan, ask, “Has anyone deliberately enabled and funded the continuation path?”

Your 30-minute check before the deadlines

Set one timer and leave with a written keep, cap, or switch decision for each tool.

Minutes 0–5: capture the live account state

Open the billing and usage screens for the accounts that fund real work. Record:

• plan name and billing owner;
• current included usage and recent consumption;
• renewal date and billing cycle;
• active paid-usage or auto-reload setting;
• the client or internal job that would pause at the limit.

Use the live screens as the account record. Public pricing pages explain the system; the account screens show which controls apply to you.

Minutes 5–15: set the Copilot boundary

For Business or Enterprise, compare recent pooled usage with the standard pool you will have after the promotion. Use assigned seats multiplied by the standard per-seat credits as the starting monthly pool, then verify the actual pool in billing.

Next, open AI Controls and inspect the paid-usage policy. Choose one boundary:

• Hard cap: disable additional paid usage. Metered features pause after the pool is exhausted, while paid-plan code completions and Next Edit suggestions remain available.
• Controlled overage: keep additional usage available and set budgets at the levels your account supports. Give heavy users an explicit limit and name the person who reviews alerts.
• Observed overage: keep the policy available for one cycle, write a dollar review threshold, name the billing owner, and set the review date now.

Convert any chosen dollar budget into credits at the official fixed rate of $0.01 per AI credit. Then write the dollar ceiling in the same place as the credit ceiling so the team can read both.

Minutes 15–25: test Claude against the standard week

Run /usage in Claude Code or open Settings → Usage. Treat the promotional 50% as temporary headroom and ask whether the standard weekly level covers the work that matters most. Keep the rolling 5-hour window as a separate constraint because the promotion leaves it unchanged.

Individual Pro and Max users should inspect usage credits. If uninterrupted work justifies pay-as-you-go spend, enable credits deliberately, set a conservative monthly limit, fund the balance, and decide whether auto-reload fits your cash controls. Team and Enterprise owners should inspect the organization-level controls shown in their live account and record the boundary that applies to their seats. If a pause is acceptable, keep the included plan limit as the boundary and schedule heavy sessions around the reset shown in the product.

Use the current cost-per-task method for any paid continuation: expected jobs multiplied by observed cost per job, compared with the value of the time or deadline protected. The live rate and your own workload produce a stronger answer than the temporary promotion.

Minutes 25–30: write keep, cap, or switch

Write one decision per service:

• Keep: standard included usage covers the recurring work, and saved time clears the current plan cost.
• Cap: the tool earns its seat, while metered agent or chat usage needs a firm dollar boundary.
• Switch: the standard allowance misses the core workload or the required overage ceiling. Move a defined class of tasks to a tested alternative and keep an exportable prompt, task, and review checklist.

For Copilot, a switch can target metered agent work while the seat continues to cover unlimited completions and Next Edit suggestions. For Claude Code, the plan can stay in place while selected overflow work moves elsewhere. Separate the workflow decision from the subscription decision.

Three questions small teams will ask

Does the GitHub promotion ending raise the seat price?

GitHub says Business and Enterprise monthly seat pricing stays unchanged. The included promotional AI-credit pool shrinks to the published standard amount. Confirm the current seat price and billing terms in your checkout because vendor pricing can change after this article’s access date.

Does Claude Code’s 5-hour limit shrink when the promotion ends?

The promotion increases the weekly Claude Code limit. Anthropic says the 5-hour usage limit keeps its existing setting, and the weekly allowance returns to standard after 31 August 2026 at 11:59 PM PT. The plan and billing continue unchanged.

Which overage setting deserves attention first?

Copilot Business and Enterprise deserve the first administrator check because additional usage starts enabled by default. Claude individual Pro and Max usage credits deserve an account-owner check because continuation requires an explicit enablement and funding flow. Team and Enterprise owners should inspect their organization settings. Record the chosen ceiling in dollars and name the person who owns the alert.

Finish the check before GitHub’s 1 September reset at 00:00:00 UTC and Claude Code’s 31 August deadline at 11:59 PM PT. A promotion can disappear cleanly while an old overage setting keeps spending, and the account control decides which outcome your team gets.