Claude Plans, Models And Usage Limits Explained (Free, Pro, Max, Ultra, Team)

10 min read

Short answer: Claude has a Free plan, Pro at $20 a month, Max at $100, Ultra at $200, and Team plans from $25 per seat. Every paid plan has a usage allowance that refills every five hours and a bigger one that refills weekly. Which model you pick, how long your chats run and how many tools you turn on decide how fast you use it. Prices below are from Anthropic's pricing page as of November 17, 2026.

When I ran my shop on Grok Bots, the thing I misunderstood most was usage. I'd hit a limit mid-job and not know why. Claude works the same way in spirit, and once I understood what actually spends the allowance, I stopped hitting walls. This guide is everything I wish someone had told me before I paid for anything.

How Much Does Claude Cost?

Plan Price Usage Notable extras
Free $0 Limited, five-hour and weekly Chat, web search, files, Artifacts, memory, connectors, up to 5 Projects
Pro $20/mo, or $17/mo billed yearly ($200) At least 5x Free per five-hour session Opus, Research, Claude Code, Claude in Chrome, scheduled tasks, Microsoft 365
Max $100/mo 5x Pro, five-hour and weekly Priority at peak times, early features, $100/mo API credits, Fable up to 50% of weekly limits
Ultra $200/mo 10x Pro weekly; no five-hour limit for verified accounts $200/mo API credits, Fable up to 50% of weekly limits
Team Standard $25/seat/mo, or $20 billed yearly More than Pro 2 to 150 users, admin controls, SSO, no training on your content by default
Team Premium $125/seat/mo, or $100 billed yearly 5x Team Standard Fable up to 50% of weekly limits

Source: claude.com/pricing. Prices exclude tax and can change. Pro is the only personal plan with yearly billing; Max and Ultra are monthly.

Which Claude Plan Should You Pick?

  • Try it out, or light personal use: Free. It's genuinely useful, and memory, connectors and Projects are included.
  • You use it most workdays: Pro. This is where Claude becomes an assistant: Opus for hard work, Research for sourced reports, Claude in Chrome for web tasks and scheduled tasks for routines.
  • You run long jobs or hit Pro's limits weekly: Max. Five times the room plus priority when Claude is busy.
  • Claude is basically your staff: Ultra. The biggest weekly allowance and, once verified, no five-hour cap.
  • Two or more people: Team. Central billing, admin controls and shared Projects, plus team features like Claude in Slack.

My honest advice: start on Pro for one month and watch Settings → Usage. If you hit the weekly wall more than once, move up. If you never get near it, you're on the right plan. A new plan's limits apply the moment you upgrade (source).

Which Claude Model Should You Use?

You choose the model in the chat box. Think of them as staff with different speeds and price tags:

Model Best for Use it when
Haiku 5.5 Speed and volume Sorting, quick lookups, short rewrites, pulling details out of text
Sonnet 5.5 Speed plus smarts Most everyday writing, replies, summaries, light research
Opus 5.5 Serious work Multi-step jobs, website or store changes, analysis, long documents
Fable 5.2 The hardest problems When Opus at high effort still falls short

Anthropic describes Opus 5.5 as the default choice for most workloads, Sonnet 5.5 as the best mix of speed and intelligence, Haiku 5.5 for high-volume quick tasks and Fable 5.2 for demanding reasoning and long jobs (source). All four can read images and use tools.

What is "effort" and "thinking"?

On current models you can also set how hard Claude thinks before answering. More thinking means better answers on tricky problems, and it also uses more allowance. Anthropic's own tip: turn thinking off or lower the effort for routine tasks (source). I keep effort low for replies and rewrites and high for anything with numbers or money in it.

How Do Claude's Usage Limits Actually Work?

There are two kinds of limit, and one length limit that people confuse with them (source):

  1. The five-hour session limit. It starts with your first message and resets in full five hours later. You can see what's left under Settings → Usage. Verified Ultra accounts don't have one.
  2. The weekly limit. A bigger allowance across all models that resets at a fixed time each week. Some plans also have a separate weekly allowance for Fable.
  3. The context window (length limit). How much one chat can hold: up to 1 million tokens on the newest models on paid plans. This isn't a usage limit; it's the size of Claude's desk.
The intricacy most people miss: every message you send, Claude rereads the whole chat so far, including files and tool results. Message 50 in a long chat costs far more than message 1. That's why long chats drain your allowance and why "one chat per task" is the most valuable habit you can build.

What uses up your allowance fastest?

  • Long chats and big attachments, because they're reread every turn.
  • Bigger models (Opus, Fable) and higher effort.
  • Tools and connectors, which Anthropic calls "token-intensive" (source).
  • Research and web searches you didn't need.

Everything shares one pool: claude.ai, the apps, Claude Code and Claude in Chrome all draw from the same allowance.

What happens when you hit a limit?

You can wait for the reset, upgrade, or, on paid plans, turn on usage credits so work continues at standard API rates instead of stopping (source). I keep a small credit cap set so a big job never stops halfway.

How Can You Make Your Claude Usage Last Longer?

  1. One chat per task. Start fresh when the job is done. Projects carry the context forward.
  2. Short Project instructions, few files. Remove files you no longer use (source).
  3. Smallest model that does the job well. Sonnet for most things; Opus when it really matters.
  4. Thinking off for routine work.
  5. Turn off connectors you aren't using in that chat, and say "don't search the web" when you don't need fresh information.
  6. Ask for the part you need. "Only the changed paragraph" costs less than "rewrite the whole thing."
  7. Batch small asks into one message instead of ten back-and-forths.

What Does Claude Cost On The API?

If you build your own tools or bots on Claude, you pay per token instead of a monthly plan. Per million tokens, input / output (source):

  • Fable 5.2: $8 / $40
  • Opus 5.5: $4 / $20
  • Sonnet 5.5: $2 / $10
  • Haiku 5.5: from $0.10 / from $0.50

All four handle up to 1 million tokens of context and up to 128,000 tokens of output. Max and Ultra include monthly API credits, which is a nice bonus if you tinker.

Your Turn

  • ☐ Open Settings → Usage and note when your session and week reset.
  • ☐ Set your default model to Sonnet; switch to Opus only for big jobs.
  • ☐ Trim your Project instructions to the rules that matter.
  • ☐ Decide on a usage credit cap you're comfortable with, or leave it off.

Frequently Asked Questions

How much is Claude Pro?

$20 a month, or $17 a month billed yearly ($200 up front) (source).

How long until Claude's usage limit resets?

The session limit resets five hours after your first message in that session. The weekly limit resets at a fixed time each week; you can see both under Settings → Usage.

What's the difference between Claude Max and Ultra?

Max is $100 a month with 5x Pro usage. Ultra is $200 a month with 10x Pro weekly usage and no five-hour limit for verified accounts.

Do Claude Code and the Claude app share usage?

Yes. The web, desktop, phone apps and Claude Code all draw from the same allowance.

Which Claude model is best?

Fable 5.2 is the most capable, Opus 5.5 is the default for serious work, Sonnet 5.5 is the fast everyday pick and Haiku 5.5 is the quickest for small jobs.

More In This Series

  1. How To Set Up Claude As Your AI Assistant
  2. Claude Plans, Models And Usage Limits Explained (you're here)
  3. How Claude Remembers: Projects, Memory, Skills And Instructions
  4. Let Claude Work In Your Apps: Connectors, Chrome And Scheduled Tasks
  5. How To Prompt Claude Like A Pro

Unofficial guide written by a Claude customer. Not affiliated with or endorsed by Anthropic. Plans and prices checked against Anthropic's pages on November 17, 2026 and may change.

Back to blog