Claude Pricing 2026: Free, Pro, Max and Team Plans Compared

Claude costs $0 on the Free plan, $20 per month on Pro (or $17 per month billed annually), $100 per month for Max 5x and $200 per month for Max 20x. Teams pay $25 per member per month for a Standard seat ($20 billed annually) or $125 for a Premium seat ($100 annually), with a two-seat minimum. Pro is the right plan for most individuals; Max is for people who hit Pro limits every day, usually because they use Claude Code heavily.

Below we explain each Claude plan as of September 2026, how Anthropic’s usage limits actually work, what Claude costs in India, and which plan fits your workload.

Claude plans and prices compared

Here is every Claude plan in one table. Anthropic describes usage as multiples of the Pro plan rather than fixed message counts, so the usage column is relative.

These figures come from the official Claude pricing page and Anthropic’s help center. If you are weighing Claude against other assistants, see our matching guides on ChatGPT pricing and Gemini pricing.

Claude Free: generous features, tight limits

The Free plan is surprisingly complete. You get chat on web, desktop and mobile, web search, file creation, code execution, memory, app connections (Anthropic calls them connectors) and Artifacts, which are interactive documents and mini apps Claude builds inside the chat.

The catch is models and volume. Free users get Sonnet and Haiku but not Opus, and usage limits are lower than Pro. Claude Code, Projects and the Chrome and Microsoft 365 extensions are paid only. The pricing page says context is “up to 1M tokens (varies by model)”.

Who it is for: students, writers testing Claude’s style, and anyone who needs an occasional long document summarized. Who should skip it: professionals who work with Claude daily, because you will hit the limit mid-task at the worst moment.

Claude Pro: $20 per month for most people

Claude Pro costs $20 per month billed monthly, or $17 per month on annual billing, paid as $200 up front. That annual option saves $40 a year compared with paying monthly, which is rarer than you might think: OpenAI has not confirmed an annual Plus price.

What Pro adds

• More usage than Free and access to Opus models, including Claude Opus 5.5.
• Claude Code, Anthropic’s coding agent for the terminal and IDE.
• Projects (see our guide on how to use Claude Projects), task scheduling, and Claude Design, Slides and Docs.
• Claude Science and the Chrome and Microsoft 365 extensions.
• Fable, available on usage credits.

How Pro limits work

Pro uses a session-based limit that resets every five hours. Anthropic’s rough estimate is “around 45 messages every five hours” for short conversations on less compute-heavy models. The real number depends on message length, attachments, how long the conversation has run and which model you use. Anthropic may also apply weekly or monthly caps, or limits on specific models or features.

Long chats cost more because Claude rereads the whole conversation each turn. Starting fresh chats and using Projects for shared context keeps you under the limit longer. Our guide to Claude usage limits and how to make them last goes deeper, and Anthropic’s own Pro plan usage article explains the rules.

Claude Max: 5x and 20x plans for heavy users

Claude Max comes in two tiers. Max 5x costs $100 per month and Max 20x costs $200 per month. They give 5x or 20x the Pro plan’s per-session usage allowance, respectively. Sessions still reset every five hours, and “Max plans also have a weekly usage limit that applies across all models”.

Anthropic’s help center estimates at least 225 to 900 messages per 5-hour session depending on tier and usage. Treat that as a guide rather than a guarantee. Max also includes Claude Code with shared limits, higher output limits, early access to new features and priority during peak traffic. Anthropic does not list an annual price for Max. Full details are on the official Max plan help page.

Who it is for: developers using Claude Code as their main coding tool, researchers processing long documents daily, and agency owners running Claude for several clients. Who should skip it: anyone who only hits the Pro limit once or twice a week. Waiting for the reset is cheaper than $80 extra per month.

Claude Team and Enterprise pricing

The Team plan works for 2 to 150 seats. You choose between two seat types and can mix them.

Both seat types include Claude Code, all models, Projects, knowledge bases, Design, Slides and Docs, and Claude Science. The pricing page describes Premium as “5x more usage than standard seats”, which matches the 1.25x and 6.25x figures.

Enterprise is priced differently: $20 per seat per month, billed annually, plus usage at API rates. That means costs scale with the model and task, so a team that uses Opus heavily pays more than one that sticks to Sonnet or Haiku. Enterprise adds advanced security and admin controls on top of everything in Team. If you want to estimate that usage, our Claude API pricing guide lists the per-million-token rates for Opus 5.5, Sonnet 5.5 and Haiku 4.5.

For help deciding between individual Pro subscriptions and a Team plan, read our explainer on when AI team plans actually save money.

Context windows and models by plan

A context window is how much text Claude can hold in mind at once, measured in tokens (about three quarters of a word each in English). On paid plans, the base chat context is 200K tokens. The newest models go further: 1M tokens on Fable 5.1, Opus 5.5, Opus 5 and Sonnet 5.5, and 500K tokens on Fable 5, Opus 4.8, 4.7, 4.6 and Sonnet 4.6. In Claude Code, most recent models support 1M tokens.

In practice, a 1M token context can hold several long reports or a sizable codebase in one conversation. Bigger context also burns through your usage limit faster, so only load what you need. Our context window explainer covers this trade-off in plain English.

Claude pricing in India

Anthropic introduced rupee pricing on July 14, 2026, according to press reports. When we checked, claude.com/pricing showed USD only, so treat these GST-inclusive figures as reported rather than official.

The Max 20x rupee price was not reported separately. Paying in rupees avoids foreign transaction fees on international cards, which can add a few percent to USD billing.

Hidden costs and how to save on Claude

Costs people miss

• Shared limits. Claude and Claude Code draw from the same allowance. A long coding session can leave you with little chat capacity for the rest of the five hours.
• Weekly caps. Max and Team seats have weekly limits across all models, and Anthropic may apply them to Pro too.
• Usage credits. Fable on Pro runs on usage credits, which are a separate spend.
• API is separate. A Pro or Max plan does not cover API calls. Developers building apps pay per token; see our prompt caching guide to cut those costs.

Ways to pay less

1. Go annual on Pro. Paying $200 up front instead of 12 x $20 saves $40 a year. Our guide to annual vs monthly AI plans explains when to commit.
2. Use the right model. Sonnet 5.5 handles most writing and analysis. Save Opus 5.5 for hard reasoning and complex code.
3. Start new chats often. Long threads cost more per message because the full history is reprocessed.
4. Mix Team seats. Give Premium seats only to heavy Claude Code users.
5. Write tighter prompts. Our library of Claude prompts for writing and editing gives you efficient starting points.

Which Claude plan should you pick?

Frequently asked questions

Verdict: which Claude plan is best value?

Claude Pro on annual billing, at $17 per month, is the best value plan for most writers, marketers and developers in 2026. It includes Opus 5.5, Claude Code and 1M token context on the newest models. Upgrade to Max 5x only when you consistently hit five-hour limits, and choose Team Standard for groups of two or more who want shared Projects and admin controls. Anthropic adjusts plans and limits regularly, so confirm current prices on the official Claude pricing page before you subscribe.

Related reading: our full Claude review and Claude vs Gemini for writing, coding and research.