ChatGPT vs Claude vs Gemini limits · Updated October 21, 2026

ChatGPT Limits: Message Caps, Token Limits & What to Do When You Hit One

ChatGPT limits come in two kinds: a usage limit on how much you can send in a time window (Plus is metered over a five-hour and a weekly window) and a length limit on how much text fits in one chat (from about 27K tokens on Free to 400K on Pro). Here is how those limits compare with Claude and Gemini — and why switching to another model is the fastest fix when you hit one.

Includes GPT-6, Claude 5.5 & Gemini 4 Argon
Use All Models in One Place See Full Comparison

Quick answer

ChatGPT, Claude and Gemini limits at a glance

Every plan has two limits: how much you can send in a time window, and how much text fits in one chat. Paid plans raise both; none is truly unlimited.

Swipe the table sideways →

App Free plan Main paid plan How usage is metered Context in the app
ChatGPT$0 · GPT-6 LunaPlus $20/monthFive-hour and weekly windows on Plus; free plan has per-feature caps27K Free · up to 256K Plus · up to 400K Pro
Claude$0 · Sonnet and HaikuPro $20/month ($17 annual); Max from $100Rolling five-hour session; weekly or monthly caps possible1M tokens on paid plans
Gemini$0 · Flash-LiteGoogle AI Pro $19.99/monthScales by plan; Ultra ($199.99) offers 20x AI Pro usage1M tokens on Google AI Pro

US prices from chatgpt.com/pricing, claude.com/pricing and gemini.google/subscriptions, checked October 21, 2026. Details for each app are below.

API Context Window at a Glance

GPT-6 Astra 1.05M tokens
Gemini 4 Argon 1,048,576 tokens
Claude Opus 5.5 1M tokens
Claude Sonnet 5.5 1M tokens
Grok 4.7 500K tokens

Scale: bar width proportional to GPT-6 Astra’s 1.05M context. Maximum API context windows from each provider’s model documentation, checked October 21, 2026. Consumer apps use smaller windows on most plans (see below).

Does ChatGPT Have a Limit?

Yes. Every ChatGPT plan has limits, including the paid ones. OpenAI’s help center describes two separate things that can stop you: a usage limit (how many messages, uploads or images you can use in a period) and a context limit (how much text the model can “see” in one conversation). When you hit a usage limit, ChatGPT shows when it resets; when a long conversation outgrows the context window, the model can no longer take all of it into account, and starting a fresh chat with a summary helps.

What are tokens?

Tokens are the basic units AI models use to process text. They aren't exactly words — they're fragments of words, punctuation marks, and whitespace that the model breaks text into before processing. As a rough guideline: 1 token ≈ 0.75 words, or conversely, 100 words ≈ 130 tokens. A 200K token context window can hold roughly 150,000 words — about the length of two full novels.

Two numbers matter: the context window (how much total text the model can "see" at once, including your prompt and the conversation history) and the max output (the longest single response the model can generate). A model with a 1M token context window but a 64K output limit can read an entire codebase but can only write back about 48,000 words per response.

ChatGPT Plus Limits (and Every Other Plan)

ChatGPT limits depend on your plan and the model you pick. On Plus ($20/month), OpenAI meters usage over a five-hour window and a weekly window; you need allowance left in both. OpenAI’s own estimates for Plus (in Work and Codex) range from about 5–45 messages per five hours on GPT-6 Astra, 15–160 on GPT-6.1 Sol and 350–3,000 on GPT-6 Luna — estimates, not guarantees, because longer prompts, files and tools use more.

ChatGPT plan Price (US) Models Message limits Context (Instant / reasoning)
Free $0 GPT-6 Luna Text chat unlimited within abuse safeguards; uploads, images, voice and data analysis have separate caps 27K / varies
Go See chatgpt.com GPT-6 Luna More messages and uploads than Free; may include ads 54K / 256K
Plus $20/month GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna and more Five-hour and weekly windows; varies by model 54K / 256K
Pro Three usage tiers All models, incl. Pro-only reasoning Unlimited within abuse guardrails; weekly allowance applies in Work & Codex 128K / 400K
Business $25/user/month ($20 billed annually); Premium seats $125 ($100) ChatGPT and Codex Premium seats: 5x Standard usage, no five-hour limit —

Sources: chatgpt.com/pricing and OpenAI help center articles on Plus, the Free tier, Business and usage limits, checked October 21, 2026. Limits change often; your Settings → Usage page shows your current allowance and reset time.

What happens when you hit the ChatGPT limit: on the free plan, GPT access pauses until the reset time shown in the app, and uploads, image creation and data analysis have their own separate caps. On Plus, Pro and Business you can wait for the reset, use a saved reset if you have one, or continue with eligible credits. One catch from OpenAI’s help center: switching to another ChatGPT model does not restore allowance in a shared usage pool — which is why many people switch to a different provider’s model instead.

Claude Usage Limits

Claude usage limits work on a rolling five-hour session. Anthropic does not publish a fixed message count; it depends on message length, attachments, how long the conversation is, and which model you use. Claude Pro ($20/month, or $17/month billed annually) gives at least five times the free plan’s usage per session — Anthropic’s own example is around 45 short messages every five hours on a less compute-intensive model. Claude Max (from $100/month) offers 5x or 20x Pro usage. Anthropic says it may also apply weekly or monthly caps.

On length, Claude is generous: on paid plans, the current chat models — Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5 — get a 1M-token context window. Usage across claude.ai, Claude Code and Claude Desktop counts toward the same limit. When you hit a usage limit you wait for the reset, upgrade, or (on Team and Enterprise) buy extra usage; when you hit a length limit you start a new chat or move the material into a Project.

Gemini Limits

In the Gemini app, Google’s plans scale by model access and usage. The free plan uses the Flash-Lite model; Google AI Plus ($4.99/month) adds the Flash model; Google AI Pro ($19.99/month) includes Gemini 4 Argon and a 1M-token context window; Google AI Ultra starts at $99.99/month, with a $199.99/month option that offers 20x higher usage limits than AI Pro.

On the API, Gemini 4 Argon has a unified 1,048,576-token context window. Its output is capped at 65,536 tokens by default and can reach the full window with long decode — useful for very long generated documents. Gemini 3.8 Flash also accepts 1,048,576 input tokens with a 65,536-token output limit.

API Token Limits Compared (October 2026)

If you build on the APIs rather than the apps, the gap has mostly closed: the flagships from OpenAI, Anthropic and Google all accept roughly a million tokens. Figures below are from each provider’s model and pricing documentation as of October 21, 2026.

Model Context Window Max Output ≈ Words ≈ Pages Input Cost / 1M Output Cost / 1M
GPT-6 Astra Flagship 1,050,000 128,000 ~790K ~3,150 $10 $50
GPT-6.1 Sol 1,050,000 128,000 ~790K ~3,150 $2 $10
GPT-6 Luna 1,050,000 128,000 ~790K ~3,150 $0.10 $0.50
Claude Fable 5.1 Flagship 1,000,000 128,000 ~750K ~3,000 $10 $50
Claude Opus 5.5 1,000,000 128,000 ~750K ~3,000 $4 $20
Claude Sonnet 5.5 1,000,000 128,000 ~750K ~3,000 $2 $10
Claude Haiku 5.5 1,000,000 128,000 ~750K ~3,000 From $0.10 From $0.50
Gemini 4 Argon Flagship 1,048,576 65,536 (up to 1,048,576 with long decode) ~786K ~3,150 $2 ($4 from Dec 1, 2026) $10 ($20 from Dec 1, 2026)
Gemini 3.8 Flash 1,048,576 65,536 ~786K ~3,150 $0.75 ($1.50 from Jan 1, 2027) $3.75 ($7.50 from Jan 1, 2027)
Grok 4.7 500,000 — ~375K ~1,500 $2 $6

Sources: OpenAI API models docs (developers.openai.com), Anthropic models overview (platform.claude.com), Google Gemini API models and pricing docs (ai.google.dev), xAI models docs (docs.x.ai). Standard per-million-token API rates; batch and caching discounts not shown. Consumer app plans have their own, smaller limits. Checked October 21, 2026.

Hit a ChatGPT Limit? Switch Models Instead of Waiting

When ChatGPT tells you to come back in a few hours, the work usually doesn’t have to stop. Different providers have separate allowances, so moving the task to Claude, Gemini or Grok keeps you going — and each model is genuinely better at some jobs. Here’s a practical guide:

Massive document analysis

Hit ChatGPT’s context limit on a 500+ page contract or report archive? Move it to a 1M-token window.

Gemini 4 Argon or Claude (1M tokens)

Codebase-level coding

Understanding and modifying entire repositories with multi-file context.

Claude Opus 5.5

Math & complex reasoning

Multi-step problems, abstract reasoning, and quantitative analysis.

GPT-6 Astra

Long-form writing

Reports, articles, essays, and any writing where quality and natural voice matter.

Claude Sonnet 5.5

Multimodal (images, video, audio)

Processing media alongside text — image analysis, video summarization, audio transcripts.

Gemini 4 Argon (native multimodal)

Not sure? Use all of them.

No single model wins everything. Send one prompt to all models and pick the best answer for each task.

MultipleChat
All Models, One Interface

Skip the Choice. Use All of Them With MultipleChat.

Message caps, context windows, pricing tiers — every model has limits. MultipleChat gives you ChatGPT, Claude, Gemini, Grok, and Perplexity in a single interface. When one model stalls, switch to another in the same chat, or send one prompt to all of them and pick the best answer. MultipleChat has its own usage limits, but you’re never stuck waiting on a single provider.

Why one subscription beats three

Access every model — ChatGPT, Claude, Gemini, Grok, and Perplexity from one place
Compare side by side — see how each model handles the same prompt differently
Auto Verification — an independent model fact-checks responses automatically
One $20 plan — MultipleChat Pro instead of separate subscriptions across several platforms

Best model for each task

ChatGPT for math, reasoning, general tasks
Claude for coding, writing, long documents
Gemini for large context, multimodal, research
All together for verification and accuracy

Frequently Asked Questions

Does ChatGPT have a limit?

Yes. Every ChatGPT plan has usage limits and a context (length) limit. On the free plan, everyday text chat is unlimited within abuse safeguards, but uploads, image creation, voice and data analysis have separate caps, and GPT access pauses until a reset when you hit the model limit. Paid plans have higher limits, not unlimited ones (Pro is unlimited within abuse guardrails).

What are the ChatGPT Plus limits?

ChatGPT Plus ($20/month) is metered over a five-hour window and a weekly window. OpenAI estimates about 5-45 messages per five hours on GPT-6 Astra, 15-160 on GPT-6.1 Sol and 350-3,000 on GPT-6 Luna (Plus in Work and Codex); these are estimates, not guarantees. The Plus context window is 54K tokens for Instant and 256K for reasoning models.

What happens when I hit the ChatGPT limit?

You can wait for the reset (the time is shown in the app or under Settings, Usage), use a saved reset if you have one, or continue with eligible credits on paid plans. Switching to another ChatGPT model does not restore allowance in a shared usage pool, so many people move the task to a different provider's model such as Claude or Gemini.

What are Claude usage limits?

Claude usage limits reset on a rolling five-hour session and depend on message length, attachments, conversation length and model. Claude Pro gives at least 5x the free plan's usage per session (Anthropic's example: around 45 short messages every five hours), and Max offers 5x or 20x Pro. Anthropic may also apply weekly or monthly caps. On paid plans, current Claude chat models have a 1M-token context window.

What is a token in AI?

A token is the basic unit AI models use to process text. Tokens aren't exactly words; they're fragments that the model breaks text into. As a rough guide, 1 token equals about 0.75 words, so 100 words is roughly 130 tokens, and 1 million tokens is about 750,000 words or 3,000 pages of text.

Which AI has the largest context window in 2026?

On the APIs it is close to a tie: GPT-6 Astra lists 1.05M tokens, Gemini 4 Argon 1,048,576 and Claude Opus 5.5 and Sonnet 5.5 1M; Grok 4.7 lists 500K. In the consumer apps, Google AI Pro (1M) and paid Claude plans (1M) offer far more than ChatGPT Plus (up to 256K) or Pro (up to 400K).

What's the difference between context window and max output?

The context window is the total amount of text (your prompt, conversation history and uploaded files) the model can see at once. Max output is the longest single response it can generate. For example, Gemini 4 Argon can read about 1M tokens but writes 65,536 tokens per response by default. These are separate limits.

Can I use all these models without paying for three subscriptions?

Yes. MultipleChat gives you ChatGPT, Claude, Gemini, Grok and Perplexity in one interface, with a free account to start and one Pro plan at $20/month. MultipleChat has its own usage limits, but when one model hits a wall you can switch to another in the same workspace, compare answers side by side and use automatic verification.

Hit a Limit? Keep Working With Another Model.

ChatGPT for reasoning. Claude for writing. Gemini for massive context. MultipleChat puts every frontier model in one place — with automatic fact-checking built in.

Try MultipleChat Free

No credit card required. Access ChatGPT, Claude, Gemini, and more in one interface.

Related resources

Compare the models beyond token limits

⚡

ChatGPT vs Gemini
🔬

Gemini vs Claude
🥊

Claude vs Grok
🏅

ChatGPT Alternatives

Use ChatGPT →

Use Claude →

Use Gemini →

Use Grok →
Read: 100 Prompts Tested — the 2026 AI Benchmark →

Continue learning

Related AI guides and next steps

See paid plans
Comparison guide

AI comparison: try the same prompt across models

Compare writing, ideas and reasoning with sample prompts and clear criteria.

 Alternatives

Best ChatGPT alternatives in 2026

See when Claude, Gemini, Grok or a multi-model workspace beats ChatGPT alone.

 Benchmark

100 prompts tested across AI models

A practical benchmark for choosing the best AI model by business task.

 Comparison

We asked 5 AIs the same question

See why serious users compare answers instead of trusting one model by default.

 Pricing

Compare MultipleChat plans

See the AI subscription that replaces separate ChatGPT, Claude, Gemini, Grok and Perplexity accounts.

 Multi-model AI

Compare AI models side by side

Run the same prompt through available ChatGPT, Claude, Gemini and Grok models and compare their answers.