Usage limits of AI coding agent plans
The monthly price is only half the story: what matters is how much work the plan lets you do before you hit a limit, and what happens then. Here is what each vendor says its plans include, in their own terms, with a link to the source. Checked Oct 9, 2026.
What about $20 a month gets you
| Plan | Included usage | Resets |
|---|---|---|
| Claude Code Pro | About 45 Claude messages or 10-40 Claude Code prompts per 5 hours | 5-hour session limit, plus weekly limits |
| OpenAI Codex Plus | Local messages per 5 hours: GPT-6.1 Sol 15-160, GPT-6 Astra 5-45, GPT-6 Luna 350-3,000 | 5-hour limit, weekly limits may apply |
| Cursor Pro | Two pools: Cursor's own models (much more included usage) and third-party models at API rates; unlimited Tab | monthly usage pools |
| Kiro Pro | 1,000 credits a month | monthly credits |
Vendors measure usage differently (prompts, messages, credits, requests), so the figures are not directly comparable. For per-token API prices see model prices.
Claude Code (Claude plans)
| Plan | Price | Included usage |
|---|---|---|
| Pro | $20/mo | About 45 Claude messages or 10-40 Claude Code prompts per 5 hours Weekly limits apply; Anthropic does not publish exact caps |
| Max 5x | $100/mo | About 225 Claude messages or 50-200 Claude Code prompts per 5 hours Weekly limits apply |
| Max 20x | $200/mo | About 900 Claude messages or 200-800 Claude Code prompts per 5 hours Weekly limits apply |
Limits reset: 5-hour session limit, plus weekly limits. When you run out: Wait for the reset, turn on extra usage (paid usage credits), upgrade, or switch to pay-as-you-go API billing.
Anthropic's own ranges; real usage depends on the model, the size of the codebase and message length. The larger models use the allowance several times faster.
OpenAI Codex (ChatGPT plans)
| Plan | Price | Included usage |
|---|---|---|
| Go | $8/mo | GPT-6 Luna in the desktop app; limits not published |
| Plus | $20/mo | Local messages per 5 hours: GPT-6.1 Sol 15-160, GPT-6 Astra 5-45, GPT-6 Luna 350-3,000 Weekly limits may apply |
| Pro | $100/mo | No 5-hour limit at present ($100, $200 or $500 a month) Weekly limits not specified |
| Business | $25/mo | Same 5-hour ranges as Plus, per user ($20 billed annually) |
Limits reset: 5-hour limit, weekly limits may apply. When you run out: Buy ChatGPT credits, or run more local sessions with an API key at API prices.
Cloud tasks use more of the allowance than local messages.
Cursor
| Plan | Price | Included usage |
|---|---|---|
| Pro | $20/mo | Two pools: Cursor's own models (much more included usage) and third-party models at API rates; unlimited Tab |
| Pro Plus | $60/mo | Same pools, larger allowance |
| Ultra | $200/mo | Same pools, largest allowance |
| Teams Standard | $40/mo | Per seat; Premium seats ($120) get 5x the Agent limits |
Limits reset: monthly usage pools. When you run out: Add on-demand usage at API rates or upgrade; requests are not downgraded.
Cursor says daily Agent users typically spend $60-100 a month and power users $200 or more.
GitHub Copilot
| Plan | Price | Included usage |
|---|---|---|
| Free | Free | A small AI-credit allowance and 2,000 code completions a month |
| Pro | $10/mo | 1,500 AI credits a month (1,000 base + 500 flex); unlimited completions |
| Pro+ | $39/mo | 7,000 AI credits a month (3,900 base + 3,100 flex) |
| Max | $100/mo | 20,000 AI credits a month (10,000 base + 10,000 flex) |
| Business | $19/mo | 1,900 AI credits per user a month |
Limits reset: monthly GitHub AI Credits. When you run out: Business and Enterprise pay $0.01 per extra AI credit; for individual plans see GitHub's billing docs.
Chat, agent mode, the cloud agent and the CLI draw on the same credits; inline completions are not billed in credits on paid plans.
Gemini CLI / Gemini Code Assist
| Plan | Price | Included usage |
|---|---|---|
| Google AI Pro | see vendor | 1,500 requests per user per day |
| Google AI Ultra | see vendor | 2,000 requests per user per day |
| Code Assist Standard | see vendor | 1,500 requests per user per day |
| Code Assist Enterprise | see vendor | 2,000 requests per user per day |
| Gemini API key (free tier) | Free | 250 requests a day, Flash model only |
Limits reset: requests per day. When you run out: Wait for the daily reset or use a pay-as-you-go API key.
Since June 18, 2026 unpaid and Google One users are moved to the Antigravity CLI. Per-minute limits also apply.
Google Antigravity
| Plan | Price | Included usage |
|---|---|---|
| Individuals | Free | Basic weekly rate limits; unlimited Tab completions and command requests |
| Google AI Pro | see vendor | Higher rate limits and a flexible AI-credit pool |
| Google AI Ultra | see vendor | Even higher rate limits and a flexible AI-credit pool |
| Business (Google Cloud) | $30/mo | Standard/Plus seats from $30 a month; pay-as-you-go option with spend caps |
Limits reset: weekly rate limits plus an AI-credit pool. When you run out: Use the credit pool or wait for the weekly reset.
Google does not publish the exact weekly limits on the pricing page.
| Plan | Price | Included usage |
|---|---|---|
| Free | Free | 50 credits a month |
| Pro | $20/mo | 1,000 credits a month |
| Pro+ | $40/mo | 2,000 credits a month |
| Pro Max | $100/mo | 5,000 credits a month |
| Power | $200/mo | 10,000 credits a month |
Limits reset: monthly credits. When you run out: Add-on credits at $0.04 each.
A one-time $20 credit when you first upgrade with social login or AWS Builder ID.
How to make a plan last longer
Most limits are spent on context the agent re-reads on every step. Clearing between tasks, smaller default models and filtering tool output stretch any plan; see how to cut token usage in Claude Code and Cursor.