Cursor vs Claude Code vs GitHub Copilot: what $10–30 a month actually buys
All three changed how they bill in 2026. GitHub Copilot moved to metered AI credits on June 1. Cursor moved its plans into the X Subscription on Oct 13. Claude renamed its top tiers on Oct 30–31. If you have $10–30 a month for an AI coding tool, here's what each one gives you right now.
The plans in the $8–30 range
| Tool & plan | Price | How usage is rationed | Models |
|---|---|---|---|
| Cursor Plus | $8/mo $80/yr | Base usage pool, refreshes weekly, shared with the Grok apps | Grok 4.7/4.6/4.5 and Composer 2.5 only, at fixed settings |
| GitHub Copilot Pro | $10/mo | $15 of AI credits a month ($10 base + $5 flex) at $0.01 per credit; completions unlimited | Choice of models; cheaper models use fewer credits |
| Claude Pro (includes Claude Code) | $20/mo $17/mo billed yearly ($200) | Usage limits per 5-hour window and per week, shared between Claude chat and Claude Code; not a fixed message count | Claude models |
| Cursor Premium | $30/mo $300/yr | 4× Plus usage, weekly; third-party models billed at API list price from prepaid credit | Grok and Composer plus third-party models; adds Cloud Agents, Automations, Bugbot, Origin |
Just above this range: Copilot Pro+ is $39/mo with $70 of credits. Claude Max is $100/mo with 5× Pro usage. Cursor Super is $100/mo with 15× Plus usage.
Three ways to ration usage
Cursor: a weekly pool
Every Cursor tier is a multiple of Plus usage (Premium 4×, Super 15×, Ultra 50×), and the pool refreshes weekly. Because it's shared with Grok Chat, Imagine, Voice and Bot, heavy use of the Grok apps eats into your coding budget.
Claude Code: 5-hour and weekly limits
Claude Pro includes Claude Code, and both share the plan's usage. Anthropic doesn't publish a fixed message count. Usage depends on the model and the task. Beyond the limits, you can buy usage credits at API rates. For reference, the API lists Sonnet 5.5 at $2 input / $10 output per million tokens and Opus 5.5 at $4 / $20.
Copilot: dollars as credits
Copilot prices usage directly in dollars: 1 credit = $0.01, and Pro includes $15 of them a month. Completions and next-edit suggestions are free on paid plans. Chat, agent mode, code review, the cloud agent and the CLI spend credits.
Same build, three tools
Test brief: [BRIEF v1 — browser game: title screen, score, restart]
What it actually cost
Test task: BRIEF v1 — small browser game (title screen, score, restart), default settings, one fresh session per tool
| Tool / plan | Sticker price | Credits used | Dollars spent | Minutes | Bugs fixed by hand |
|---|---|---|---|---|---|
| Cursor Plus | $8/mo | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] |
| GitHub Copilot Pro | $10/mo | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] |
| Claude Code Claude Pro | $20/mo | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] |
| Cursor Premium | $30/mo | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] | [TEST RESULT PENDING] |
Scorecard
Which to pick (pricing only, not yet tested)
- Tightest budget: Cursor Plus is the cheapest at $8, if Grok-only models are fine for you. Is Cursor Plus worth it?
- Mostly autocomplete: Copilot Pro's completions are unlimited and don't spend credits. Is GitHub Copilot Pro worth it?
- Terminal-first agent work: Claude Pro includes Claude Code. How far its limits stretch: [TEST RESULT PENDING]
- Want Cursor's full feature set: Premium at $30 adds third-party models, Cloud Agents, Automations and Bugbot, none of which come with Plus.
Verdict pending
[TEST RESULT PENDING] — no verdict until the hands-on test is done. We don't guess.
Sources
Prices and plan details checked Nov 1, 2026. Prices change often; if you spot a change, tell us and we'll update.
- Cursor — Pricing
- Cursor Docs — Models & pricing
- Claude — Pricing — Pro/Max/Ultra, Claude Code included, API prices
- GitHub Copilot — Plans & pricing
- LimitDex — secondary: Claude Max/Ultra rename on Oct 30–31, 2026
- PlanDiff — secondary: Cursor move to X Subscription on Oct 13, 2026; Copilot metered billing June 1, 2026