How Many Tokens Do You Get with Claude Pro and Max?
Anthropic doesn’t publish a token allowance. How Claude’s session and weekly limits work, Pro vs Max vs Ultra, and what each plan’s price buys on the API.
By Polarison · Updated November 11, 2026
Short answer: Anthropic doesn’t publish a token allowance for Claude Pro or Max. Subscription limits are a usage budget that shrinks faster with long conversations, big files, tools and the most capable models. What you can pin down is what each plan’s price would buy on the API, which tells you whether the subscription is a good deal for how you work.
How Claude’s subscription limits actually work
According to Anthropic’s help center, usage limits are a “conversation budget” over a period of time. How fast you use it depends on:
- the length of your messages and of the conversation so far,
- file attachments and tools such as Research and web search,
- the model you choose and the effort level,
- multi-step tasks like running code or creating files.
Usage in claude.ai, the desktop app and Claude Code all counts against the same limit. The session limit resets every five hours on Pro and Max, and every plan also has a weekly limit across all models, which resets at a fixed time assigned to your account. Ultra has no five-hour limit for verified accounts, only the weekly one.
Pro vs Max vs Ultra
- Pro ($20 a month) is the baseline allowance, with a five-hour session limit and a weekly limit.
- Max ($100 a month) gives five times Pro’s usage every five hours and every week, and includes Fable up to half the weekly limit.
- Ultra ($200 a month) gives ten times Pro’s weekly usage and, for verified accounts, has no five-hour limit.
Ultra replaced the earlier Max 20x plan at the same $200 price. The difference that matters most for coding is the missing five-hour wall: long Claude Code sessions don’t pause mid-day, though the weekly cap still applies.
What each plan’s price buys on the API
Here is the same money spent on API tokens instead, using a typical message of 4,000 input tokens (prompt plus conversation so far) and 800 output tokens.
| Plan | Price | Claude Opus 5.5 messages | Claude Sonnet 5.5 messages | ≈ tokens on Claude Sonnet 5.5 |
|---|---|---|---|---|
| Claude Pro | $20.00 | 625 | 1,250 | 6.0M |
| Claude Max | $100.00 | 3,125 | 6,250 | 30.0M |
| Claude Ultra | $200.00 | 6,250 | 12,500 | 60.0M |
If you would send more than this in a month, and stay inside the plan’s session and weekly limits, the subscription is cheaper than the API. Try your own numbers in the subscription vs API calculator.
How big can a single conversation be?
That is a separate limit: the context window. On paid plans, Anthropic says the newest models support up to a 1M token context window, while others support 500K or 200K tokens. A 1M-token conversation is roughly 870,000 words. Long chats are summarized automatically when code execution is on, but that summarizing uses more of your usage limit.
Ways to make your limit last longer
- Start a new conversation when the topic changes: every message resends the whole history.
- Put reference files in a Project. Project content is cached and counts less when reused.
- Batch related questions into one message instead of several.
- Use Claude Sonnet 5.5 for routine work and save Claude Opus 5.5 for hard problems; the model choice changes how fast you use the allowance.
When you hit the limit
You can wait for the reset, buy usage credits to keep going, upgrade to a Max tier, or switch Claude Code to a pay-as-you-go Console account. One trap: if an ANTHROPIC_API_KEY environment variable is set, Claude Code bills the API instead of your subscription. Compare plans on the Claude pricing page and API rates on Claude API pricing.