Search

Full-text search

Changes ledger

Every recorded change to a tracked plan's price or limits, with the baseline stated. Entries are never edited silently; corrections are logged.

Last verified 17 days old, May be outdated. Flagged as outdated after 9 days. AI-assisted check  11 changes, 2 in the past month · How we verify

1 change

  1. GitHub Copilot No net change Minor

    Copilot Business and Enterprise seats are charged upfront for credit card and PayPal customers

    From October 1, 2026, GitHub charges existing Copilot Business and Copilot Enterprise customers who pay by credit card or PayPal upfront for all assigned seats at the start of each billing cycle. For these customers, each new seat must also be paid for before its user gets Copilot access. GitHub says the prices are not changing.

    BeforeSeats were billed each month for the number assigned, with billed users counted at the end of each billing cycle. Credit card and PayPal payers paid upfront only for seats added during a cycle, prorated for the rest of that cycle.

    AfterFor credit card and PayPal payers, all assigned seats are charged upfront at the start of each billing cycle, and each new seat is paid for before its user gets access. Seats added mid-cycle are still prorated, and included usage may be prorated to match. Prices unchanged: Business $19 and Enterprise $39 per user per month.

    Baseline: Seat billing before October 1, 2026, as GitHub Docs describe it. On October 2, 2026 the Docs still described billed seats counted at the end of each cycle and did not yet mention the start-of-cycle charge. Prices are compared like for like: no change. Also not changing, per GitHub: no prorated refund when a seat is revoked, and proration for seats added mid-cycle.

    Sources: GitHub Changelog: Upcoming changes to GitHub Copilot policies and billing, GitHub Docs: GitHub Copilot seats and billing cycles for organizations and enterprises , verified 17 days agoMay be outdatedAI-assisted check

2 changes

  1. Codex Mixed

    ChatGPT Pro: a new $500 plan, Pro 200 reopens with a lower allowance, no five-hour limit for Pro currently

    On September 29, 2026 OpenAI introduced Pro 500, a $500/month ChatGPT Pro plan with the highest included usage of the Pro plans and access to Astra Ultrafast, and made Pro 200 available for new subscriptions again at $200/month. New Pro 200 subscriptions that are not eligible for grandfathering get a lower usage allowance than before; eligible subscribers keep the previous allowance through October 29, 2026. OpenAI now names the Pro plans Pro 100, Pro 200 and Pro 500, and its Codex pricing page says Pro plans currently have no five-hour limit.

    BeforePro was sold as a $100 tier and a $200 tier, described as 5x or 20x more Codex usage than Plus, with five-hour message estimates listed for both; new sign-ups and upgrades to the $200 plan had been paused since September 10, 2026.

    AfterPro 100 ($100), Pro 200 ($200) and Pro 500 ($500) are on sale, and only Pro 500 includes Astra Ultrafast. New Pro 200 subscriptions that are not eligible for grandfathering get a lower allowance, and eligible existing ones move to it after October 29, 2026. Pro plans currently have no five-hour limit; weekly limits may also apply.

    Baseline: Compared with the Pro tiers of September 29, 2026: the $100 tier introduced as the 5X tier and the $200 plan that OpenAI called Pro 20X. The $100 and $200 prices did not change. OpenAI no longer states multiples of Plus, gives no size for the lower Pro 200 allowance and lists no Pro five-hour estimates, so the change in usage is not quantified here. The page does not say when the five-hour limit disappeared: the copy saved on September 29, 2026 still listed five-hour estimates for Pro, and the copy of October 2, 2026 says Pro plans have no five-hour limit.

    Sources: OpenAI Help Center (OpenAI): ChatGPT — Release Notes, OpenAI Help Center (OpenAI): About ChatGPT Pro tiers, OpenAI: Pricing | ChatGPT Learn, OpenAI: Pricing | ChatGPT Learn (copy of 29 Sep 2026, before the Pro plans were restructured), OpenAI Help Center (OpenAI): About ChatGPT Pro tiers (copy of 19 Sep 2026, while Pro 200 sign-ups were paused), VentureBeat: OpenAI introduces ChatGPT Pro $100 tier with 5X usage limits for Codex compared to Plus , verified 17 days agoMay be outdatedAI-assisted check

  2. Claude Decrease

    Claude Code weekly limits settle at 25% above the pre-promotion level

    A promotion that raised Claude Code weekly limits by 50% ran through September 13, 2026. From September 14, 2026, weekly limits in Claude Code are permanently 25% higher than before the promotion, for Pro, Max, Team and seat-based Enterprise plans. Anthropic described this as a 17% reduction compared with the promotional level.

    Limit ledger · what changed
    1. Before the promotionPre-promotion level100
    2. Promotion, now endedThrough Sep 13, 2026150
    3. Permanent levelFrom Sep 14, 2026125Limit
    Hatched: at or above today's limit. An index computed by LimitSheet, not absolute values. Anthropic publishes no absolute weekly values.
    Official: Stated by the vendor on its own pricing page, docs or help center. Claude Help CenterReported: Reported by the vendor elsewhere (blog, staff post) or by reputable press. BleepingComputer, verified 17 days agoMay be outdatedAI-assisted checkThe change of

    BeforeWeekly Claude Code limits at the promotional level, 50% above the pre-promotion level (through September 13, 2026).

    AfterWeekly Claude Code limits 25% above the pre-promotion level, permanently, for Pro, Max, Team and seat-based Enterprise.

    Baseline: Two baselines. Index the pre-promotion weekly limit at 100: the promotion made it 150 and the new level is 125. Against the promotional level: 125 / 150 − 1 = −16.7% (Anthropic: "a 17% reduction in weekly limits on Claude Code"). Against the original pre-promotion level: 125 / 100 − 1 = +25%. Anthropic publishes no absolute weekly values.

    Sources: Claude Help Center (Anthropic): Claude Code May–August 2026 weekly limits promotion, BleepingComputer (Mayank Parmar): Anthropic is cutting Claude Code's current weekly limits by 17% , verified 17 days agoMay be outdatedAI-assisted check

2 changes

  1. Cursor Plan or price added

    Cursor launches Start, a rupee-billed plan for India

    Cursor launched Start, a lower-priced plan for developers in India, billed monthly in rupees with UPI or card; Cursor's docs describe the price as tax inclusive. Start covers the Cursor Models pool and Cloud Agents, but not the Other Models pool or on-demand usage.

    BeforeIndividual plans: Free (Hobby) and Pro and above, with no plan priced in rupees.

    AfterStart added for India: billed in rupees, tax inclusive, monthly billing only, UPI or card. Other plans unchanged.

    Baseline: The individual plan lineup before the launch. No existing plan changed; a plan was added. The rupee price is the vendor's own, not an FX conversion.

    Sources: Cursor: Introducing Cursor Start, Cursor: Models & Pricing | Cursor Docs, Cursor: Cursor · Cursor Start , verified 17 days agoMay be outdatedAI-assisted check

  2. Claude Plan or price added

    Claude starts showing INR prices in India (reported)

    TechCrunch reported that local rupee pricing had begun to appear for some users in India on Claude's website and mobile apps, with local taxes included. Payment was by card or app store billing; UPI was not enabled at the time of the report.

    BeforeClaude plans priced in US dollars for users in India.

    AfterTechCrunch reported local rupee prices for Pro, Max and Team, taxes included. This site shows a rupee price only after a person has seen the amount at a real checkout (see the India page).

    Baseline: Dollar pricing for users in India, as the same report describes it. LimitSheet never converts US dollars to rupees. The date is the report date, not a confirmed rollout date.

    Source: TechCrunch (Jagmeet Singh): Anthropic starts localizing Claude pricing for India, its biggest market after the US , verified 17 days agoMay be outdatedAI-assisted check

2 changes

  1. GitHub Copilot Mixed

    GitHub Copilot moves from premium requests to GitHub AI Credits

    GitHub replaced premium request units (PRUs) with GitHub AI Credits on all Copilot plans. Usage is now calculated from token consumption at each model's listed API rates. Plan prices did not change. Monthly Pro and Pro+ plans moved on June 1, 2026; annual Pro and Pro+ plans stay on premium requests until they expire.

    BeforePremium requests (PRUs): each plan included a monthly premium request allowance, with model multipliers.

    AfterGitHub AI Credits ($0.01 each), based on input, output and cached tokens at listed API rates. Prices unchanged: Pro $10 a month, Pro+ $39, Business $19 per user, Enterprise $39 per user. Code completions and next edit suggestions do not use credits.

    Baseline: Premium-request billing in effect before June 1, 2026, announced for replacement on April 27, 2026. Prices are compared like for like (no change). Included usage is not directly comparable before and after, because premium requests and AI Credits measure different things. Copilot code review also started consuming GitHub Actions minutes.

    Sources: The GitHub Blog: GitHub Copilot is moving to usage-based billing, GitHub: GitHub Copilot · Plans & pricing, GitHub Docs: Usage-based billing for individuals , verified 20 days agoMay be outdatedAI-assisted check

  2. GitHub Copilot Increase

    GitHub Copilot adds flex allotments to Pro and Pro+ and a new Max plan

    Announced May 12, 2026 and effective June 1, 2026: Pro and Pro+ include more total usage at the same price through a variable flex allotment on top of base credits, and a new Max plan was added at $100 a month.

    BeforeAs first announced in GitHub's usage-based billing post: Pro $10 a month including $10 in monthly AI Credits; Pro+ $39 including $39; no Max plan.

    AfterPro $10 with $15 of included usage ($10 base + $5 flex); Pro+ $39 with $70 ($39 + $31); new Max $100 with $200 ($100 + $100). In credits: 1,500, 7,000 and 20,000 a month.

    Baseline: The included AI Credits in GitHub's first usage-based billing announcement, which equaled the plan price. Base credits stay matched 1:1 with the price. GitHub says “Flex allotments will vary over time”, so later flex changes will be logged as their own entries.

    Sources: The GitHub Blog: GitHub Copilot individual plans: Introducing flex allotments in Pro and Pro+, and a new Max plan, The GitHub Blog: GitHub Copilot is moving to usage-based billing, GitHub Docs: Plans for GitHub Copilot , verified 20 days agoMay be outdatedAI-assisted check

2 changes

  1. Claude Increase

    Claude Code weekly limits raised 50% in a limited-time promotion

    From May 13, 2026 through September 13, 2026, weekly usage limits in Claude Code were 50% higher for Pro, Max and Team plans and for legacy seat-based users on Enterprise plans. Free plans and consumption-based Enterprise seats were not included. Five-hour limits and the limits of other Claude products were not affected.

    BeforeStandard weekly Claude Code limits (the pre-promotion level).

    AfterWeekly Claude Code limits 50% above the pre-promotion level through September 13, 2026, for Pro, Max, Team and legacy seat-based Enterprise users.

    Baseline: The standard weekly Claude Code limits in effect before May 13, 2026. Anthropic publishes no absolute weekly values, so the change is stated only as a percentage of that level. The article's title says May–August, but its text says the promotion ran through September 13, 2026. The level that followed from September 14, 2026 is a separate entry. The cited source does not say when the promotion was announced.

    Source: Claude Help Center (Anthropic): Claude Code May–August 2026 weekly limits promotion , verified 20 days agoMay be outdatedAI-assisted check

  2. Claude Increase

    Claude Code five-hour limits doubled for paid plans

    Anthropic doubled Claude Code's five-hour rate limits for Pro, Max, Team and seat-based Enterprise plans, and removed the peak-hours limit reduction on Claude Code for Pro and Max accounts.

    BeforeClaude Code five-hour limits at the previous level; a peak-hours limit reduction applied to Claude Code on Pro and Max.

    AfterClaude Code five-hour limits doubled for Pro, Max, Team and seat-based Enterprise; no peak-hours reduction on Pro and Max.

    Baseline: The five-hour limits in effect just before the announcement. Anthropic published no absolute values, so the change is stated only as a multiple of that baseline (doubled). The announcement does not mention weekly limits.

    Source: Anthropic: Higher usage limits for Claude and a compute deal with SpaceX , verified 20 days agoMay be outdatedAI-assisted check

1 change

  1. Codex Plan or price added

    OpenAI adds a $100 ChatGPT Pro tier with 5x Plus Codex usage

    OpenAI launched a $100 ChatGPT Pro tier with five times the Codex usage of Plus, next to the $200 plan, then called Pro 20x. At launch, the limits shown for Pro 5x included a temporary 2x usage boost that was set to end May 31, 2026.

    BeforeOne ChatGPT Pro plan, at $200 a month.

    AfterTwo Pro tiers: Pro 5x at $100 a month (5x Plus Codex usage) and Pro 20x at $200 a month (20x Plus).

    Baseline: The ChatGPT plan lineup just before the launch, as reported by VentureBeat. No existing plan changed price; a tier was added. OpenAI's pricing page, in the saved copy cited here, defined the Pro tiers relative to Plus (5x and 20x); the per-model message figures it published were estimates, not fixed message limits. OpenAI's Pro tiers article now lists the $100 and $200 Pro plans as Pro 100 and Pro 200, the names the plan tables on this site use.

    Sources: VentureBeat: OpenAI introduces ChatGPT Pro $100 tier with 5X usage limits for Codex compared to Plus, OpenAI: Pricing | ChatGPT Learn (copy of 29 Sep 2026, before the Pro plans were restructured), OpenAI Help Center (OpenAI): About ChatGPT Pro tiers , verified 17 days agoMay be outdatedAI-assisted check

1 change

  1. Claude Decrease

    Claude adds weekly limits for Pro and Max

    Anthropic added two weekly rate limits for Pro and Max subscribers on top of the existing limits that reset every five hours: an overall weekly limit and a separate weekly limit for Claude Opus 4. Both reset every seven days.

    BeforePro and Max usage limits that reset every five hours; no weekly limit.

    AfterFive-hour limits unchanged, plus an overall weekly limit and a weekly Claude Opus 4 limit, each resetting every seven days.

    Baseline: The limits in place before August 28, 2025, which reset every five hours and had no weekly cap. Anthropic published no absolute weekly values. It estimated the new limits would apply to less than 5% of subscribers, and told TechCrunch that most Pro users could expect 40 to 80 hours of Sonnet 4 through Claude Code within their weekly limits. Announced July 28, 2025.

    Source: TechCrunch (Maxwell Zeff): Anthropic unveils new rate limits to curb Claude Code power users , verified 20 days agoMay be outdatedAI-assisted check

Atom feed

How the ledger works

  • An entry records one change: what it was before, what it is now, and the baseline the comparison uses.
  • We describe changes plainly (“reduced”, “raised”) and always name the baseline, because a change can be a cut against one level and a rise against another.
  • Entries are append-only. A mistake gets a dated correction on the entry and in the corrections log.
  • Follow new entries with the Atom feed, or download all entries from the data page.