Skip to content
Claude News

Guide

Claude prices and limits

What Claude plans and the API cost and what their limits are

Updated 21 October 202684 stories

In short

Current API list prices: Opus 5.5 costs $4 input and $20 output per million tokens, with cache reads at $0.20 (source). Sonnet 5.5 costs $2/$10 per Mtok with 1M context (source), and its cache reads have since been halved from $0.20 to $0.10 per million tokens (source). Haiku 5.5 costs $0.10 input and $0.50 output per million tokens on prompts under 100k (source). Opus 5.5 is the default on paid plans. Because it is priced below Opus 5, 5-hour and weekly limits go 25% further (source).

Subscription limits: since September 14, Claude Code weekly limits have been a permanent 25% above the original baseline. Compared with the temporary 50% boost that ended September 13, that is a 17% drop (source). Five-hour session limits rose 20% on September 23, and the spare reset in Settings → Usage can be applied until Oct 22 (source). On Max 20x, the 20x applies to the 5-hour limits, while the weekly allowance works out at 10x (source). Anthropic doesn’t publish how many tokens Max 5x or Max 20x buys (source).

What changed most recently: in October, Max 5x started getting $100 a month in API credits and Max 20x $200. Team plans earn $20 per Standard seat and $100 per Premium seat, with the shared balance capped at $500. Pro gets nothing, and the credits don’t cover interactive Claude Code sessions or extra usage (source). Claude Code 2.1.304 now asks before you resume an inactive conversation on Pro and Max, because a resume can use a noticeable share of the 5-hour limit (source). The startup offer of a free Team year plus a $1,000 API credit is paused (source).

At a glance

Plan or modelWhatValueAs ofSource
Claude Opus 5.5API price$4 input and $20 output per million tokens; cache reads $0.20 per million tokens22 September 2026Read
Claude Opus 5.5Default in Claude CodeDefault model on Pro and Team Standard plans from Claude Code 2.1.280, with a 1M context window22 September 2026Read
Claude Sonnet 5.5API price and context$2/$10 per Mtok, 1M context28 September 2026Read
Claude Sonnet 5.5Cache reads$0.10 per million tokens, down from $0.207 October 2026Read
Claude Haiku 5.5API price$0.10 input and $0.50 output per million tokens on prompts under 100k; smaller cut on prompts over 100k7 October 2026Read
Claude Fable 5.1Cache reads$0.25 per million tokens, down 75%; rest of pricing unchanged from Fable 51 September 2026Read
Pro, Max, Team, seat-based EnterpriseClaude Code weekly limitsPermanent 25% increase over the standard cap from September 1429 August 2026Read
5-hour session limitsChangeRose 20%23 September 2026Read
Max 20xLimit multiplier20x applies to the 5-hour limits; weekly allowance works out at 10x31 August 2026Read
Max 5xMonthly API credits$100 a month7 October 2026Read
Max 20xMonthly API credits$200 a month7 October 2026Read
TeamMonthly API credits$20 per Standard seat, $100 per Premium seat; one shared balance capped at $5007 October 2026Read
Show all (17)
Plan or modelWhatValueAs ofSource
ProMonthly API creditsNone7 October 2026Read
Claude Code wrap-up at 5-hour limitAvailabilityPro: once a week; Max and Team Premium: each time a 5-hour session limit is hit25 September 2026Read
Claude Team plan for scientistsPrice and seatsNo cost for groups of principal investigators at academic and nonprofit research institutions; 10,000 seats; Premium seats with 5x usage limits $15 per month for one year28 August 2026Read
Claude StartupsFree Team year and $1,000 API creditPaused; Anthropic will re-review every application12 October 2026Read
Requests blocked by safeguardsBillingCharged again in biology, distillation attacks and frontier LLM development24 September 2026Read

Milestones

  1. Anthropic boosts Claude Code limits by 50 percent.
  2. Anthropic releases the Claude Fable 5 and Mythos 5 models.
  3. Anthropic faces a lawsuit over subscription limits.
  4. Claude Fable 5 becomes a permanent part of Max and Team Premium.
  5. Claude Opus 5 launches at Opus 4.8 pricing and becomes the Max default.
  6. Claude Fable starts requiring usage credits.
  7. Sonnet 5’s $2/$10 intro pricing becomes permanent.
  8. Anthropic announces a permanent 25% increase to weekly limits to replace the temporary 50% boost.
  9. Fable 5.1 ships with cache reads cut 75% to $0.25 per million tokens.
  10. Opus 5.5 launches at $4/$20 per million tokens, down from $5/$25 on Opus 5.
  11. Five-hour session limits rise 20%, and paid users get a spare limit reset.
  12. Claude Code cloud sessions leave research preview, with a one-time credit for Pro and Max.
  13. Anthropic resumes billing for requests its safeguards block in three categories.
  14. Sonnet 5.5 becomes the default Sonnet on the API at $2/$10 per Mtok with 1M context.
  15. Claude Startups expands its offer of a free Team year plus $1,000 in API credits.
  16. Haiku 5.5 launches at $0.10/$0.50 per million tokens on short prompts, and Sonnet 5.5 cache reads are halved.
  17. Max and Team plans get monthly API credits that don’t cover Claude Code sessions.
  18. The free Team year and API credit for startups are paused after a flood of applications.

FAQ

How much does Claude Opus 5.5 cost on the API?

$4 per million input tokens and $20 per million output tokens, with cache reads at $0.20 per million. Opus 5 cost $5 and $25. Anthropic puts the saving at 40% on typical workloads (source).

How much do Claude Sonnet 5.5 and Haiku 5.5 cost?

Sonnet 5.5 costs $2/$10 per Mtok with 1M context (source). Its cache reads were later halved from $0.20 to $0.10 per million tokens (source).

On prompts under 100k, Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens. Prompts over 100k get a smaller cut (source).

What are the current Claude Code weekly limits?

Since September 14, standard weekly limits have been a permanent 25% above the original baseline on Pro, Max, Team and seat-based Enterprise plans. The temporary 50% boost ended September 13, so compared with that boost, subscribers get 17% less (source). Five-hour session limits rose 20% on September 23 (source).

Does Claude Max 20x really give 20 times the usage of Pro?

The 20x applies to the 5-hour limits. The weekly allowance works out at 10x (source). A class action says the separate weekly ceiling wasn’t adequately disclosed at signup for the $200-a-month tier (source).

Do Claude Max and Team plans include API credits?

Yes. Max 5x gets $100 a month and Max 20x gets $200. Team plans earn $20 per Standard seat and $100 per Premium seat, all in one shared balance capped at $500. Pro gets nothing. The credits don’t cover interactive Claude Code sessions or extra usage, and they show up once you’ve been on an eligible plan for seven days (source).

Latest