Why now
Cursor's pricing page was rebuilt in October 2026. It now shows Free, Plus, Premium, Super and Ultra under "The X Subscription" and no longer mentions Pro or Pro+ (see ToolPulse Weekly #4). Our earlier comparison, GitHub Copilot vs. Cursor vs. Replit: How Their AI Credits Work, still describes the old Cursor ladder, so this piece replaces it as the current reference. GitHub Copilot also has a new top individual tier, Max, since that earlier piece was written.
Pricing at a glance
As of Nov 2, 2026, per each vendor's official pricing page (individual plans; USD, excluding tax).
| Tool | Plan | Price | What the plan includes |
|---|---|---|---|
| Cursor | Plus | $8/mo | Basic usage; Grok models only |
| Cursor | Premium | $30/mo | 4x Plus usage; third-party models billed extra |
| Cursor | Super | $100/mo | 15x Plus usage; priority at peak times |
| Cursor | Ultra | $200/mo | 50x Plus usage |
| GitHub Copilot | Pro | $10/mo | 1,500 AI credits ($15 value) |
| GitHub Copilot | Pro+ | $39/mo | 7,000 AI credits ($70 value) |
| GitHub Copilot | Max | $100/mo | 20,000 AI credits ($200 value) |
| Replit | Core | $18/mo, billed annually | $20 towards more powerful models |
| Replit | Pro | $90/mo, billed annually | $100 credits; up to 15 collaborators |
Sources: cursor.com/pricing, cursor.com/docs/account/pricing, github.com/features/copilot/plans, docs.github.com — usage-based billing for individuals, replit.com/pricing, all retrieved Nov 2, 2026. Free tiers, team plans and Enterprise are covered in the sections below, not in the table. Replit's page shows $20 and $18 side by side for Core and $100 and $90 for Pro, labelled "billed annually"; we read the lower figure as the annual-billing price and do not state a separate monthly-billing price. Cursor's docs list annual billing of $80/yr (Plus), $300/yr (Premium), $1,000/yr (Super) and $2,000/yr (Ultra).
What differs for a working developer
Cursor
Cursor's plans now differ mainly by a multiple of the Plus allowance: Premium is 4x, Super 15x and Ultra 50x. The included pool covers Cursor models and is shared across the editor, CLI, cloud agents and the Grok apps. According to Cursor's docs it refreshes weekly, so "one heavy day doesn't use up your month".
The key change is third-party models. Per the docs, Claude, GPT and Gemini sit outside the weekly pool. On Premium and above they are billed at API list price from the first request, paid from prepaid credit. Plus and Free do not include them at all. Users can top up manually or turn on Auto Top-up, and set a monthly cap on additional usage. When the weekly pool runs out, the docs list three options: wait for the reset, use prepaid credit, or upgrade. They add that requests "are never downgraded in quality or speed".
Two further points from the docs. Teams seats cost $30, $100 or $200 per month depending on the tier (the pricing page says "$30+"), and eligible third-party requests on Teams and Enterprise carry an extra "Cursor Token Rate" of $0.25 per million tokens that individual plans do not pay. And the previously unconfirmed question about existing customers now has an official answer: Cursor's docs say Pro, Pro+, legacy Ultra, Start and Teams seats "keep working as they are", with two monthly usage pools and on-demand usage. The docs do not list prices for those legacy plans. The pricing page lists the legal seller as X Subscription Corp. and shows Anysphere, Inc. as copyright holder; it does not explain the relationship.
GitHub Copilot
Copilot has the most explicit credit math. Per GitHub, "1 AI credit = $0.01 USD". Each paid individual plan gets base credits equal to its price in dollars, plus a variable "flex" allotment: Pro 1,000 + 500, Pro+ 3,900 + 3,100, Max 10,000 + 10,000. GitHub says flex amounts are variable and may change as model pricing evolves. Base credits are used first.
Credits are consumed by Copilot Chat, CLI, the cloud agent, Spaces, Spark and third-party coding agents. Code completions and next edit suggestions are not billed in credits and stay unlimited on paid plans. The Free plan has no credits; the pricing page limits it to 2,000 completions and entry-level models per month, with limited chat and agent usage.
When credits run out on an individual plan, GitHub lists three options: upgrade and pay the difference, stay on the plan and set a dollar budget for extra usage, or wait. Allowances reset at 00:00 UTC on the first day of each month and unused credits are forfeited. The plans page does not list Business or Enterprise prices; it says admins on those plans set usage limits and decide whether extra paid usage is allowed. We therefore leave organization pricing out of this comparison.
Replit
Replit's model is the simplest to state: the plan price is the credit amount. Core lists "$20 towards more powerful models"; Pro lists "$100 credits" and "$100 towards most powerful models". Pro allows up to 15 collaborators and offers more Free Mode usage. The pricing page shows no separate free plan, though Core mentions limits for "Free Mode" (up to 30 hours of chat and up to 60 projects).
Replit's billing documentation covers what happens at the limit. Usage limits "cap your spending beyond your monthly credits", and when a limit is reached, "usage-based services are blocked until the next billing cycle or until you increase the limit". Core and Pro users can also buy credit packs of $100, $300 ($290), $500 ($480) or $1,000 ($950); packs expire six months after purchase and do not renew automatically. The docs we read do not say whether unused monthly subscription credits roll over, so we make no claim either way. The pricing page also shows higher credit levels ($215, $425, $825 and $2,000 per month) without labelling them clearly, so we do not describe them.
What differs for budgeting
- Is the model bill inside the price? Copilot and Replit: the plan price gives a credit balance that pays for model use. Cursor Premium and above: only a Cursor-model pool is included; third-party models are a separate prepaid bill.
- How is usage metered? Cursor: a weekly pool measured as a multiple of Plus. Copilot: monthly credits at $0.01. Replit: monthly dollar credits.
- Is there a hard stop? Replit: yes, once limits are reached. Copilot: your choice of upgrade, budget or wait. Cursor: wait, prepaid credit or upgrade.
- What is flat? Only Copilot states that completions stay unlimited on paid plans.
Who should pick what
- Wants the lowest entry price and is fine with Grok models: Cursor Plus at $8/mo. It does not include Claude, GPT or Gemini.
- Wants Claude, GPT or Gemini inside a fixed monthly price: check GitHub Copilot or Replit first. On Cursor you pay API list price for those models on top, so total cost depends on your usage.
- Wants a published exchange rate between dollars and usage: GitHub Copilot ($0.01 per credit, base and flex spelled out).
- Wants a built-in spending ceiling: Replit's usage limits block usage at the limit; Cursor and Copilot both offer user-set caps or budgets.
- Is an existing Cursor Pro or Pro+ subscriber: per Cursor's docs, your plan keeps working as is. Check your billing page before switching.
What this comparison doesn't cover
We do not estimate real monthly costs: they depend on the model chosen, project size and how much agent or cloud-agent work you run, and none of the pages quantify that. We also do not compare code quality, speed or editor experience, and we leave out Business, Enterprise and team pricing where the vendors do not publish it on the pages cited. Some Cursor models carry promotional pricing; the docs say one runs "through November 21, 2026".
Sources & methodology
This comparison is based on the vendors' public information as of Nov 2, 2026, not on hands-on testing.
- Cursor — Pricing, retrieved Nov 2, 2026
- Cursor Docs — Account & pricing, retrieved Nov 2, 2026
- Cursor — Changelog, retrieved Nov 2, 2026 (no pricing entry found)
- GitHub — Copilot plans, retrieved Nov 2, 2026
- GitHub Docs — Usage-based billing for individuals, retrieved Nov 2, 2026
- Replit — Pricing, retrieved Nov 2, 2026
- Replit Docs — Managing your spend, retrieved Nov 2, 2026