How many messages can I send on ChatGPT Plus?

On ChatGPT Plus you get a much larger GPT-5 allowance than Free — around 160 messages in a rolling ~3-hour window, after which GPT-5 auto-switches to GPT-5 mini until the window resets. Reasoning models like GPT-5 Thinking carry a separate weekly ceiling. OpenAI tunes these numbers by server load, so treat them as a guide, not a contract.

When does the Claude Pro limit reset?

Claude Pro uses a rolling 5-hour window: the counter clears five hours after your first counted message, not at a fixed clock time. If your first message lands at 09:00, the window resets around 14:00. Pro also has a weekly cap for very heavy use, which resets seven days after that week's usage started.

What happens when you hit the ChatGPT message limit?

On Free and Plus, once you exhaust the GPT-5 quota, ChatGPT auto-switches you to a lighter model (GPT-5 mini) for the rest of the window instead of stopping outright. You keep chatting, just on a smaller model, until the rolling window resets and the top model is available again.

Does Claude have a weekly limit as well as a 5-hour limit?

Yes. Anthropic added weekly limits in 2025 on top of the rolling 5-hour window, aimed at the heaviest users. So two caps apply: a short-window cap (about 45 messages every 5 hours on Pro) and a weekly cap. Whichever you reach first is the one that stops you — this calculator checks both.

Is ChatGPT Plus worth it for the number of messages I send?

If your usage regularly pushes past the Free GPT-5 cap and you dislike being downgraded to a mini model, Plus at about $20/month (roughly Rs 6,000 at a 300 rate) removes that wall for typical daily use. If you only occasionally hit the limit, the free fallback model may be enough. The upgrade panel above prices the break-even for your exact usage in rupees.

How does this calculator estimate my limit?

It divides your messages-per-session by your active hours to get a per-hour rate, multiplies by each plan's window length to get messages-per-window, and compares that to the published cap. It also projects your weekly total against any weekly cap. The reference caps are cited constants; the windowing and reset math are fully deterministic.

Are these message limits exact?

No, and no third-party tool can promise exact numbers. Vendors adjust caps often, and a limit stated as messages can shift with message length, attachments, and server load. The values here are each provider's published or publicly-guided figures with a visible verification date. Always confirm against the official help page for anything mission-critical.

Does Gemini or Perplexity limit reset daily instead of hourly?

Often, yes. Google Gemini and Perplexity typically meter their premium modes (Gemini 2.5 Pro, Perplexity Pro search) on a daily basis rather than a short rolling window, then fall back to a lighter mode once you hit the daily count. This calculator marks those as fixed-window resets so the reset clock reflects a daily boundary, not a 5-hour one.

AI · Usage limits

AI Chat Message Limit Calculator — ChatGPT, Claude & Gemini

Find out how many messages you get before hitting your plan's cap on ChatGPT, Claude, Gemini, Grok, or Perplexity — when the window resets, which model you drop to at the wall, and whether upgrading is worth it in rupees. No signup, sources cited.

By Induwara Ashinsana— Executive Director, Ryzera TechnologiesUpdated Jul 4, 2026

Message limit check

Limits verified 2026-07-04

Provider

Plan

Primary model

Messages per session

How many you send in one sitting (1–2000).

Active hours / day

Hours you actively chat (0.5–16).

Days per week

Days you use it (1–7).

Window start time

When your first message lands — used for the reset clock.

USD → LKR rate

Editable. Used only for the rupee upgrade prices.

Usage presets

You'll hit the 5-hour limit after 45 messages — about 15 spill over and must wait until 14:00.

Your window load

over each 5 h window

Plan cap

messages / 5 h

Spill over the cap

wait or downgrade

Window resets

14:00

rolling window

What happens at the wall

At the cap, Claude Sonnet 4.5 stops responding (hard stop) until the window resets.

Weekly check

300 messages/week vs a 900/week cap — within the weekly limit.

Upgrade recommendation

Max (5×) buys you 180 extra messages per window for about $100/month.

Plan

Max (5×)

Monthly

$100.00

In rupees

Rs 30,500

New cap / window

225

Caps, windows and prices are each provider's published or publicly-guided figures — they change often and vary by server load and message length. Last verified 2026-07-04. Full sources are listed below the calculator.

How it works

Every AI chat plan meters usage inside a window— a rolling stretch of a few hours, or a fixed daily boundary. The calculator turns your everyday habits into a messages-per-window figure and compares it to the plan's published cap. All caps, window lengths, weekly limits, fallback models, and prices are cited constants stored in one data module with a 2026-07-04 verification date; the arithmetic below is fully deterministic.

Usage rate. messages per hour = messages per session ÷ active hours per day. This assumes your session is spread across your active hours; if you send them in a burst, you hit the cap sooner, so this is the conservative reading.
Messages per window. window messages = messages per hour × window length. For a Claude Pro 5-hour window, 12 messages/hour becomes 60 per window.
Rolling-limit test. If window messages exceed the cap C, the limit is hit: the spill-over is window messages − C. Landing exactly on the cap counts as within the limit — the test is strictly greater than, so there is no off-by-one at the boundary.
Reset clock. A rolling window clears the window length after your first counted message, so a 09:00 start on a 5-hour window resets at 14:00. Fixed-window providers (Gemini 2.5 Pro, Perplexity Pro search) reset on a daily boundary instead.
Weekly test. Where a plan has a weekly cap Cw (for example Claude Pro and Max), weekly messages = messages per session × days per weekis checked against it. Whichever cap you reach first — rolling or weekly — is reported as the real blocker.
Fallback behaviour. If the model has a fallback (GPT-5 → GPT-5 mini), the tool tells you what you are downgraded to at the wall; otherwise it reports a hard stop until reset.
Upgrade break-even.If you exceed a cap, the tool scans the same provider's higher plans in ascending price and picks the first whose same-tier model clears both your window and weekly demand, converting the monthly price to rupees at price in USD × your rate.

As a self-check, the window-messages figure is computed a second way — (messages per session × window length) ÷ active hours — which is algebraically identical, so the two paths always agree to the message.

Worked examples

Claude Pro, heavy debugging session

Cap 45 msgs / 5 h window · weekly cap 900 · you send 60 over 5 active hours, 5 days/week

Messages per hour: 60 ÷ 5 = 12
Window messages: 12 × 5 = 60
60 > 45 cap → limit hit; spill-over 60 − 45 = 15
Reset: 09:00 first message + 5 h = 14:00
Weekly: 60 × 5 = 300 ≤ 900 → weekly cap clear
Verdict: 15 messages wait until 14:00; Max (5×) raises the cap to ~225/window

ChatGPT Free, deciding on an upgrade

GPT-5 cap 10 / 5 h window · fallback GPT-5 mini · you send 30 over 6 active hours

Messages per hour: 30 ÷ 6 = 5
Window messages: 5 × 5 = 25
25 > 10 cap → after 10 GPT-5 messages, auto-switch to GPT-5 mini for the other 15
Upgrade scan: Plus GPT-5 cap 160 ≥ 25 → fits
Rupee price: $20 × 300 = Rs 6,000/month
Verdict: Plus keeps GPT-5 for all 25 messages for about Rs 6,000/month

Edge case — landing exactly on the cap

Claude Pro cap 45 / 5 h window · you send 45 over 5 active hours

Messages per hour: 45 ÷ 5 = 9
Window messages: 9 × 5 = 45
45 > 45 is false → NOT hit; you sit exactly on the cap
Headroom left: 0 — one more message tips you over
Verdict: within the limit, but with no room to spare

Frequently asked questions

Sources & references

Provider caps, windows, and prices were last cross-checked against these help pages on 2026-07-04. Vendors change these figures often and vary them by server load and message length, so confirm against the official page before relying on an exact number.

Related tools

LiveAI

AI Plan Comparison

Compare the paid consumer plans of ChatGPT, Claude, Gemini, Grok and Perplexity side by side — message limits, models unlocked, vision, voice, file upload, image generation and deep research, plus price in USD and Sri Lankan rupees. Enter your usage and get the cheapest plan that fits. Sources cited, no signup.

Open tool

LiveAI

AI Max Output Tokens

Look up the maximum output (completion) tokens for every current LLM — Claude, GPT-4o, Gemini, Llama and more — and check whether your desired response fits in a single API call or needs chunking. Per-model caps cited from vendor docs, separate from the context window.

Open tool

LiveAI

AI Video Token Cost Calc

Estimate how many input tokens a video costs when you send it into a multimodal LLM — Gemini's native per-second tokenization versus frame-sampling into GPT-4o and Claude — priced per video and per month in USD and LKR. Runs in your browser; no video is uploaded.

Open tool

Rate this tool

Be the first to rate

Comments & feedback

Spotted a bug or want an improvement? Tell us — our team reviews every comment, and good ideas get built. Comments are public and anonymous.

Spotted a limit that's changed, or want another provider added?

Email me at [email protected] — most fixes ship within 24 hours.