RoninForge / Copilot Bill Migrator
GitHub Copilot bill migrator
On June 1, 2026 every Copilot plan moved from premium requests to token-metered AI credits. Enter what you used under the old model and see how your monthly bill changes, and whether it now costs you more. Everything runs in your browser.
Legacy allowances and the $0.04 overage verified against GitHub's docs on 2026-06-12.
Your old usage
How that usage prices in AI credits
Your bill, before and after
- Old bill (premium requests)
- $30.00
- $10.00 plan + $20.00 overage
(500 requests over 300 at $0.04) - New bill (AI credits)
- $52.24
- $10.00 plan + $42.24 overage
(4,224 credits over 1,500 at $0.01)
At this usage you draw about 5,724 credits/month, past the Copilot Pro allowance of 1,500.
Old model vs new model, by plan
Same subscription price, different metering. Premium requests were a flat count with $0.04 overage; AI credits are token-metered with $0.01 per credit beyond the allowance.
| Plan | Price | Old: premium requests | Old overage | New: AI credits | New overage |
|---|---|---|---|---|---|
| Copilot Pro | $10.00/mo | 300/mo | $0.04/req | 1,500/mo | $0.01/credit |
| Copilot Pro+ | $39.00/mo | 1,500/mo | $0.04/req | 7,000/mo | $0.01/credit |
| Copilot Business | $19.00/seat/mo | 300/mo | $0.04/req | 1,900/mo | $0.01/credit |
| Copilot Enterprise | $39.00/seat/mo | 1,000/mo | $0.04/req | 3,900/mo | $0.01/credit |
Copilot Max ($100/mo, 20,000 credits) is new with usage-based billing and had no legacy premium-request tier. Business and Enterprise credits pool at the billing entity, with promotional amounts until September 1, 2026. Code completions stay unlimited on paid plans under both models.
How this estimate works
The old bill is exact: your subscription plus $0.04 for every premium request beyond your plan's included allowance.
The new bill is an estimate. Premium requests count interactions; AI credits price the tokens inside them, so there is no fixed conversion. We take your premium-request count as a rough interaction count, price one interaction in credits from the official per-model token rates and your chosen session size, and bill any credits beyond the new allowance at $0.01 each.
This is most accurate when your premium requests went mostly to 1x-multiplier models. If you leaned on high-multiplier models like Opus or the o-series, each interaction counted as several premium requests, so your real interaction count was lower and this estimate runs high. Lower your premium-request number, or pick the model and session size that match your work, to tighten it.
Frequently asked questions
Did my GitHub Copilot bill go up on June 1, 2026?
It depends on your model mix and session size. The June 1 switch moved every monthly plan from premium requests (a flat count with $0.04 overage) to AI credits (token-metered, 1 credit = $0.01). Light chat users on small models often pay about the same or less. Heavy agent-mode users on powerful models tend to pay more, because long contexts and large generations burn many credits per interaction. This tool estimates the direction and size of the change for your usage.
What is the difference between a premium request and an AI credit?
A premium request was one billed interaction, counted against a fixed monthly allowance (300 on Pro, 1,500 on Pro+), with $0.04 per request beyond it. An AI credit is a unit of token spend: GitHub prices each interaction by its input, cached, and output tokens at per-model rates, converts the dollar amount to credits at $0.01 each, and draws from your plan allowance (1,500 on Pro, 7,000 on Pro+, 20,000 on Max). Requests measure count; credits measure tokens.
How do I find how many premium requests I used?
On the legacy usage page in your GitHub billing settings, your premium-request total for recent months is listed. Enter that number above. If you only know it roughly, estimate: the comparison is directional, and you can move the slider to see how sensitive your bill is.
Is one premium request the same as one AI credit?
No. They are different units, so there is no fixed conversion. This estimator treats one premium request as roughly one model interaction, then prices that interaction in credits from the token rates. If you leaned on high-multiplier models like Opus or the o-series, each of those counted as several premium requests, so your real interaction count was lower and this estimate runs high. Pick your main model and session size above to tighten it.
I am on an annual Copilot plan. Did anything change for me?
Annual Pro and Pro+ subscribers who stayed on legacy request-based billing keep premium requests until their annual term ends, then move to AI credits (or to Copilot Free) unless they change plans. All monthly plans moved to AI credits on June 1, 2026.
Is this affiliated with GitHub?
No. This is an independent community tool from RoninForge. It runs entirely in your browser, sends nothing you enter anywhere, and produces estimates only. Your actual bill is computed by GitHub from your real token usage.
Sources
- GitHub Docs: Requests in GitHub Copilot (legacy)
- GitHub Docs: Overview of request-based billing (legacy)
- GitHub Docs: Models and pricing for GitHub Copilot
- GitHub Blog: Copilot is moving to usage-based billing
Independent community tool, not affiliated with GitHub. Estimates only; your bill is computed by GitHub from your real token usage. Rates change as models change, so check the sources for current numbers.
Copilot AI Credits calculator
Estimate your monthly credit burn from token usage, exactly, and compare it against your plan allowance.
Open calculatorCopilot usage CSV analyzer
Drop the usage report GitHub emails you and see where the credits went: by model, by day, by SKU. Runs fully in your browser.
Analyze your report
Copilot is one line of your AI spend. Track every provider in one dashboard, with history, per-project attribution, and budget alerts. See the dashboard
Want your whole stack in one number, subscriptions included? Total it with the AI stack cost auditor.