Understand your Claude usage. Stretch it further.
One page for people who use Claude privately and for teams that run it at work. Every number links to its official source and shows when we last checked it.
Four things that decide how far you get
Session limit
Your short-term allowance. It resets every five hours.
Weekly limit
A longer allowance that resets once a week at a fixed time assigned to your account.
Length limit
The context window: how much one conversation can hold. It is separate from usage limits.
Usage credits
Optional extra usage you can buy so work continues after you hit a limit.
What applies to you
Private
The session limit resets every five hours.
Weekly limits reset once a week, at a fixed time assigned to your account. Your next reset time is shown in Settings > Usage.
Usage limits are shared across claude.ai, Claude Code and the desktop app. Work in one place uses the same allowance as the others.
Usage limits and length limits are different things. Length is the context window of a single conversation.
Max 5x gives five times the Pro per-session usage. Max 20x gives twenty times. Max also has a weekly limit across all models.
On Pro and Max you can enable usage credits to keep working after you reach a limit.
Automatic context management (compaction) consumes additional usage.
Business
Team Standard seats get 1.25 times the Pro per-session usage. Premium seats get 6.25 times. Both have a weekly limit across all models.
Team usage is per member, not pooled. One person reaching a limit does not affect the others, and limits cannot be transferred.
Team admins can enable usage credits so members keep working after they hit their limits.
Claude Code is included with every Team seat. Premium seats include more usage.
Team monthly API credits cannot be used in Claude Code or the Claude apps.
Usage-based Enterprise has no per-seat usage limits. The seat fee covers access only, all usage is billed at API rates, and admins can set spend caps.
Legacy seat-based Enterprise plans keep their Standard and Premium allowances until they are migrated.
Across enterprise deployments, Anthropic reports an average of about 13 USD per developer per active day for Claude Code, typically 150 to 250 USD per month, with 90 percent of users under 30 USD per active day. Your own cost will differ.
See where your usage goes
/usageShows session estimate, plan limit bars and a breakdown. The dollar figure is an estimate; the Console is authoritative for billing./contextShows what currently occupies your context window./clearStarts a fresh conversation. It costs nothing and is the cheapest way to switch topic.
Habits that save usage
Use /clear when the topic changes. It is free, and it stops a long history from being carried into every new message.
Thinking tokens are billed as output tokens, so heavy reasoning settings cost more.
MCP tool definitions are deferred by default in Claude Code, so unused tools do not fill your context.
After a long idle period the prompt cache is cold, so the next message re-reads the whole context at full cost.
Agent teams use far more tokens than a single session. The docs cite roughly seven times in plan mode.
Switching to a different model does not raise your session or weekly limit. It only helps if you hit a model-specific limit.
What we cannot tell you
Better to say so than to guess.
- The exact day and hour of your weekly reset. It is fixed per account; Settings > Usage shows yours.
- How many messages or tokens a session limit equals. This site gives no fixed number, because we found none in the official sources; usage depends on message length, files, model and features.
- How much of the weekly limit any one model uses. Check /usage in Claude Code or Settings > Usage.