Skip to content
StackVersus

Data report · checked 2026-11-06

AI plan limits report, November 2026

What 74 plans across 13 AI assistants and coding agents (Claude, ChatGPT + Codex, Cursor, GitHub Copilot, Google AI (Gemini, Antigravity), Devin (Windsurf), Replit, Lovable, v0, Bolt, Kiro, Amp, Augment Code) actually promise. Every figure comes from the vendor's own page, with the source and the date we read it, on the limits directory.

22 / 74

plans state no number for any usage limit: they describe it only in words, or not at all

38%

of the 359 limits listed carry a number (a count, a dollar amount or a multiple of another plan)

12

plans stack a five-hour window on top of a weekly cap

6

tools sell an individual plan at $200 a month or more: Claude, ChatGPT + Codex, Cursor, Google AI (Gemini, Antigravity), Devin (Windsurf), Amp

What plans meter

Plans by the unit their usage is counted in

  • Allowance, pool or pay-as-you-go29

  • Credits29

  • Tokens or dollars7

  • Messages5

  • Runner minutes4

When limits reset

Plans with at least one limit on each window (a plan can have several)

  • Monthly39

  • Weekly22

  • Five-hour window20

  • Daily8

What they cost

Plans by monthly list price (62 with a public price)

  • Free14

  • $1–2417

  • $25–9914

  • $100–1999

  • $200 and up8

Plans that state no number

These plans describe their limits only in words or leave them out. Buyers cannot work out in advance how much use the price buys.

Changes since 2026-05-07

15 recorded changes to prices, metering or limits across 5 tools, newest first. Full log: /changes.

  1. 2026-11-06 · Kiro: metering rules rewritten (current wording on the tool page).Source
  2. 2026-11-06 · Kiro added the Teams (credit pooling) plan at $20 per seat a month. Add-on credits: $0.04 per credit, bought by admins for the pool; expire 12 months after purchase; Credits: 500 included per seat per month, shared by all seat holders in the pool; Overage: None; credit pooling doesn't use overages; Rollover: Unused included credits expire at the end of the monthly cycle; Usage controls: Admins set credit limits for individuals, groups and profiles; alerts at 80% and 100% utilization.Source
  3. 2026-11-01 · ChatGPT + Codex: new limit "Shared allowance from December 2026": Starting in December, usage allowance is shared across ChatGPT and Codex and measured in tokens; Instant chat does not draw from it. OpenAI expects most users won't wait for a reset more often than now. Plans: Business, Plus, Pro ($100), Pro ($200), Pro ($500).Source
  4. 2026-10-31 · Claude Pro: Session usage changed from At least 5x the Free per-session allowance to More usage than Free every 5 hours; no multiple statedSource
  5. 2026-10-31 · Claude Max: name changed from Max 5x to MaxSource
  6. 2026-10-31 · Claude Max: Session usage changed from 5x the Pro per-session allowance to 5x the Pro plan's usage every five hoursSource
  7. 2026-10-31 · Claude Max: Weekly usage changed from Weekly limit across all models; separate weekly limit for Fable to 5x the Pro plan's usage every week; separate weekly limit for FableSource
  8. 2026-10-31 · Claude Ultra: name changed from Max 20x to UltraSource
  9. 2026-10-31 · Claude Ultra: Session usage changed from 20x the Pro per-session allowance to No five-hour limit for verified accountsSource
  10. 2026-10-31 · Claude Ultra: Weekly usage changed from Weekly limit across all models; separate weekly limit for Fable to 10x the Pro plan's usage every week; separate weekly limit for FableSource
  11. 2026-10-30 · Amp added the Education (Megawatt at 50% off) plan at $10 a month.Source
  12. 2026-10-30 · Bolt Free: Web requests changed from 333K per month to 333,333 per month (hard limit)Source
  13. 2026-10-30 · Bolt added the Lite plan at $9 a month.Source
  14. 2026-10-30 · Bolt Pro: annualMonthlyUsd is now $18Source
  15. 2026-10-30 · Bolt Pro: Tokens changed from 10M per month, no daily limit to Starts at 10M per month (10M on the $25 tier)Source

Using these numbers

Free to quote and republish under CC BY 4.0 with a link to this page. Suggested citation:

StackVersus, "AI plan limits report, November 2026", https://getstackversus.com/limits/report (data checked 2026-11-06)

Method: we read each vendor's pricing, help and documentation pages, copy the wording, and record a limit as numeric only when the vendor states the number. The page regenerates whenever the data is re-checked. Raw data: changes JSON, daily GitHub mirror. Press questions: contact.