Which plan should you pick?
Claude Code comes with a Claude subscription. You don’t buy it separately. The same plan also covers Claude on the web, desktop and mobile, and they all draw on one shared allowance.
- Free doesn’t include Claude Code.
- Pro ($20/month, or $17/month billed yearly) includes Claude Code everywhere it runs: terminal, VS Code and JetBrains, the desktop app and the web. You get Opus 5.5 (the default model), Sonnet 5.5 and Haiku 5.5, with at least five times Free’s usage per session. It’s enough for most part-time or moderate daily use. Claude Fable isn’t in Pro’s limits and runs on usage credits only.
- Max 5x ($100/month) gives five times Pro’s usage per session. It’s the usual upgrade once Pro’s limits start interrupting your work.
- Max 20x ($200/month) gives twenty times Pro’s usage, for people who run Claude Code most of the day or keep several sessions going. On both Max plans, Fable can use up to half of the weekly limits.
- An API key instead of a plan suits automation, CI jobs and irregular bursts, where paying per token beats a monthly fee. See paying per token below.
Max is billed monthly only. Annual billing exists only for Pro and Team, and Anthropic says there are no standing discounts on Pro or Max.
How the usage limits work
Claude’s paid plans have two limits:
- A rolling 5-hour session. Your allowance resets five hours after your session starts.
- A weekly limit, which resets at a fixed time assigned to your account.
Anthropic says there’s “no fixed message count”. It doesn’t publish hours or prompts per plan, and the figures you’ll see quoted from 2025 aren’t on any current official page. What you get depends on the model, the length of your conversations and the size of your codebase. On September 14, 2026 Anthropic permanently raised weekly Claude Code limits by 25%.
When you hit a limit, you can:
- wait for the session or weekly reset;
- upgrade to Max, or to a Premium seat on Team;
- turn on usage credits (Anthropic’s new name for “extra usage”). These are prepaid, billed at standard API rates, with an optional monthly cap and auto-reload. They’re available on Pro, Max, Team and seat-based Enterprise;
- switch Claude Code to an API key and pay per token.
One surprise to avoid: if the ANTHROPIC_API_KEY environment variable is set and you approve it, Claude Code bills that key at API rates instead of using your plan.
Paying per token with an API key
You can run Claude Code without a subscription. Use an API key from the Claude Console, or run it through Amazon Bedrock, Google Cloud’s Agent Platform or Microsoft Foundry. You pay per token at these rates:
| Model | Input | Output | Cache write (5 min) | Cache read |
|---|---|---|---|---|
| Opus 5.5 (Claude Code’s default) | $4 | $20 | $5 | $0.20 |
| Sonnet 5.5 | $2 | $10 | $2.50 | $0.10 |
| Haiku 5.5, prompts up to 100K tokens | $0.10 | $0.50 | $0.125 | $0.01 |
| Haiku 5.5, prompts over 100K tokens | $0.50 | $2.50 | $0.625 | $0.05 |
| Fable 5.1 | $10 | $50 | $12.50 | $0.25 |
Prices are US dollars per million tokens. One-hour cache writes cost twice the input price, US-only inference costs 1.1 times as much, and Opus 5.5 fast mode is $8 input and $40 output.
How much is that in practice? Anthropic’s Claude Code docs say: “Across enterprise deployments, the average cost is around $13 per developer per active day and $150-250 per developer per month, with costs remaining below $30 per active day for 90% of users.” Older blog posts quote $6 a day; that figure is out of date.
For steady daily use, $150–250 a month is more than Pro and about the same as Max 20x. So a subscription usually wins if its limits fit your work. Claude Code’s /cost command and the Console’s usage page show what you’re spending.
What the API credits in Max and Team cover
Since October 7, 2026, Max includes $100 or $200 a month in API credits. Team gets pooled credits worth $20 per Standard seat and $100 per Premium seat, capped at $500 a month. These credits don’t pay for interactive Claude Code or for usage credits. They cover the Claude API, the Agent SDK, Managed Agents, and claude -p scripts run with an API key from a linked Console organization.
To claim them, you link a Console organization after seven days on the plan. Unused credits don’t roll over. Pro, Free and Enterprise don’t get them.
Team and Enterprise pricing
Team is for 2 to 150 people. Every seat now includes Claude Code. Standard seats got it on January 16, 2026, and seat prices were cut on January 28, 2026.
| Seat | Monthly | Billed yearly | Usage per session |
|---|---|---|---|
| Standard | $25 per seat | $20 per seat | 1.25x Pro |
| Premium | $125 per seat | $100 per seat | 6.25x Pro |
Limits are per member. Owners can prepay usage credits for the team and cap them per organization or per user. Admins get spend limits for the organization and for each user, and can export a spend report.
Enterprise costs $20 per seat per month, billed annually, plus all usage at API rates. There are no per-seat usage limits, so usage credits don’t apply. The minimum is 20 seats if you buy self-serve (available since February 12, 2026), where you prepay credits. It’s 50 seats through sales, which bills monthly in arrears. Enterprise adds group-level spend limits and an analytics API.
Discounts
- Annual billing: Pro ($200 a year, or about $17 a month) and Team only.
- Nonprofits: discounted Team and Enterprise. Anthropic doesn’t publish the rates.
- Research groups: a verified program gives free Standard seats and $15 Premium seats on Team for 12 months.
- Education: “discounted rates” on request, with no published prices.
- Startups: Anthropic’s startup program offers up to $100K in API credits through partner VCs. Its current cohort is full.
Cursor
OpenAI Codex
GitHub Copilot
Gemini CLI
Kiro