Claude Pricing 2026: Free, Pro, Max, Ultra and API

Last updated: November 2026. Claude's plans and models have both moved a lot since April, per Anthropic's pricing page and API pricing docs:

• Ultra is a new $200/month plan. The $200 slot used to be "Max 20x". Max is now a single $100 plan.
• Max and Ultra include monthly API credits: $100 on Max, $200 on Ultra.
• Fable, Anthropic's top model tier, is included on Max and Ultra for up to half your weekly limit. On Pro it is available through paid usage credits. Current model line-up: Fable 5.2 (Nov 9), Opus 5.5 (Sep 22), Sonnet 5.5 (Sep 28) and Haiku 5.5 (Oct 7).
• API prices came down. Opus 5.5 is $4/$20 per million tokens (input/output) versus $15/$75 for the Opus 4 we listed in April, and Sonnet 5.5 is $2/$10.
• Claude Code is now included in every paid plan, and the product family has grown (Claude Design, Cowork, Science, Slides, Docs and more).

Claude is now a full product family: consumer plans, team plans, enterprise contracts and a pay-per-token API. For individuals there are four plans (Free, Pro, Max, Ultra), plus Team and Enterprise for organizations.

Prices are as of November 2026 and exclude tax. Check the official Claude pricing page before you buy, since Anthropic notes that prices and plans can change.

The table gives you the numbers. Here is what each tier delivers and where the gaps are.

Free Plan ($0/month)

Free covers web, desktop and mobile chat, web search, file creation and code execution, memory, Artifacts, connectors, and Claude Design, Slides and Docs, with limited usage every 5 hours and every week. You get Sonnet and Haiku but not Opus, and no Claude Code. It is enough to evaluate Claude, not to depend on it.

Pro Plan ($17/month annual, $20 monthly)

Pro is where Claude becomes a work tool. It adds Claude Code, Claude Science, Projects, Research, handoff and scheduled tasks, Claude Data plus Sheets and Whiteboards (beta), Claude in Chrome and Microsoft 365, and access to Opus. Fable, the top tier, is available through paid usage credits rather than being included. Anthropic says Pro gives at least 5x more usage per 5-hour session than Free. Paying annually costs $200 upfront.

Max Plan ($100/month)

Max is a single $100 plan now. You get 5x Pro usage every 5 hours and every week, Fable up to half of your weekly limit, early access to advanced features, priority access at busy times and $100 a month in API credits for the Claude Platform. For developers who live in Claude Code, the API credits alone offset a good part of the price if you also build on the API.

Ultra Plan ($200/month)

Ultra includes everything in Max, with 10x Pro usage every week, no 5-hour limit for verified accounts, and $200 a month in API credits. It is for people whose work is long, uninterrupted agent sessions. It replaces what used to be sold as Max 20x at the same price (as we described it in April).

Team Plan

Team is for 2 to 150 seats. Standard seats are $25 monthly or $20 on annual billing and give more usage than Pro. Premium seats give 5x Standard and cost $125 monthly or $100 annual. Teams get up to $500 a month of pooled API credits, central billing and admin controls.

Enterprise Plan

Anthropic lists Enterprise at $20 per seat per month, billed annually, plus usage at API rates, so the bill scales with the models and tasks your team runs. It includes everything in Team plus SCIM, audit logs, custom data retention and role-based access. Anthropic notes you can start self-serve or contact sales for a tailored quote.

API pricing is separate from the consumer plans (except for the monthly credits on Max, Ultra and Team). Anthropic charges per token with different rates for input and output. Per its pricing docs, the current line-up as of November 2026 is:

Claude Mythos 5.2 shares the same underlying model and price as Fable 5.2 but has fewer safeguards and is limited to vetted Project Glasswing partners for defensive cybersecurity. Older models are still listed, such as Opus 4.5 to 5 at $5/$25, Sonnet 4.6 at $3/$15 and Haiku 4.5 at $1/$5.

• Prompt caching: cache hits cost a fraction of base input (for example $0.20 per 1M on Opus 5.5 and $0.10 on Sonnet 5.5).
• Tokenizer change: Anthropic notes Claude 4.7 and later models produce about 30% more tokens for the same text, so compare real bills, not just price per token.
• US-only inference (inference_geo "us") costs 1.1x, and fast mode is available on some models at premium prices.
• Batch API discounts and extended thinking: check the docs for current batch rates; thinking tokens are billed as output.

For most applications, start with Opus 5.5 as Anthropic recommends, drop to Sonnet 5.5 where quality holds, and use Haiku 5.5 for classification, extraction and summarization at scale. Reserve Fable for the hardest work. See the official pricing page for the latest.

Claude Pro costs $20 monthly ($17 on annual billing), in line with ChatGPT Plus. The real question is whether it delivers enough over Free and over its competitors.

What Pro Actually Changes

The jump from Free to Pro unlocks Opus, Claude Code, Research, Projects, scheduled tasks and the Chrome and Microsoft 365 integrations, with a lot more usage. If you routinely hit the Free limit, Pro removes that bottleneck. One caveat: the top Fable model is not included on Pro and runs on usage credits, so if you want it regularly you are looking at Max.

Who Should Subscribe to Pro

• Writers and content creators: Claude's writing is a favourite for long-form, nuanced content.
• Developers: Claude Code is included. See our Claude Code skills guide for what is possible.
• Researchers and analysts: models support up to 1M-token context, and Pro includes Research.
• Anyone who hit Free limits more than twice: simple math.

Who Should Skip Pro

If you use Claude a few times a week for quick questions, Free is fine. If you mainly need real-time search, Perplexity may serve you better. And if you want the widest feature set in one subscription, ChatGPT covers more ground.

Claude and ChatGPT now have almost mirror-image ladders at the top. Here is how they line up as of November 2026, from each vendor's pricing page.

Where Claude wins: coding with Claude Code, long documents, and writing that does not sound robotic. Where ChatGPT wins: breadth, with image generation, voice, the always-on dots agent on Pro, and a cheaper Go tier.

The honest take: many power users pay for both. If you must pick one, document-heavy and code-heavy work favors Claude; diverse, multimedia workflows favor ChatGPT. For a broader look see our AI writing tools and AI coding tools comparisons.

If you are building products, API cost matters more than consumer pricing. Here is a snapshot per million tokens as of November 2026, using standard rates from each vendor's page or our shared fact sheet:

• Sonnet 5.5 vs GPT-6.1 Sol is the like-for-like mid-tier comparison and list prices are identical at $2/$10. Quality, not price, decides it.
• Budget tier: Haiku 5.5 and GPT-6 Luna both start at $0.10 input / $0.50 output, with Haiku rising above 100K-token prompts.
• Claude's tokenizer produces about 30% more tokens for the same text than older models, so measure on your own data.
• Gemini 4 Argon and Carbon carry introductory pricing that Google has said ends soon, so we left them out of the table.

For automation workflows, run a benchmark on your actual data before committing. The cheapest model on paper is not always cheapest once you count retries and output quality.

Not sure which plan fits? Here is a use-case-driven guide.

For Students and Learners

Start with Free, upgrade to Pro during crunch periods. Free handles homework help and feedback. For exams or a thesis, Pro at $20 is worth the temporary upgrade.

For Writers and Content Creators

Pro is the sweet spot. Consider Max if you produce at volume. Browse our AI writing tools directory for how Claude stacks up against specialized writing tools.

For Developers and Engineers

Pro for most. Max or Ultra for those running Claude Code all day. Read our Claude Code skills and MCP guide for the full breakdown. Max and Ultra also include API credits, which is a real perk if you build on the platform. Also explore our AI coding tools page.

For Small Business Teams

Team Standard at $25/seat monthly ($20 annual). Admin controls and central billing matter the moment you have 2+ people. If you also want to automate marketing workflows or e-commerce operations, pair Team with the API.

For Researchers and Analysts

Pro minimum, Max if your work is document-heavy. Models support up to 1M-token context, and Max gives 5x the usage so you do not hit walls on a heavy research day.

Here are practical ways to reduce your Claude costs.

Consumer Plan Tips

• Use the right model for the task. Don't default to the biggest model for everything. Lighter models handle many tasks and stretch your limits.
• Use Projects to reduce repetition. Store standing instructions and context instead of re-explaining every chat.
• Pay annually if you stay. Pro is $17/month on annual billing ($200 upfront) versus $20 monthly.
• Compare Max to dual subscriptions. If Claude covers most of your needs, Max at $100 with $100 of API credits may beat paying for several $20 plans, especially if you also use the API.

API Cost Optimization

• Enable prompt caching. Cache hits are billed at a small fraction of base input.
• Use the Batch API for async work. Check the pricing docs for the current batch discount.
• Route by model. Send complex tasks to Opus or Fable, standard tasks to Sonnet, and high-volume simple tasks to Haiku.
• Minimize output tokens. Output costs 5x input across current models. Ask for concise, structured responses.

Claude's ladder is easier to read than it used to be: Free, Pro at $20, Max at $100 and Ultra at $200. Pro is still the best-value tier for most individuals. Max now makes sense for heavy daily users, and its $100 API credit softens the price for developers. Ultra is for the longest, most intensive sessions. Team and Enterprise add governance.

On the API side, the current generation is cheaper than the models we covered in April, with Opus 5.5 at $4/$20 and Sonnet 5.5 at $2/$10. Anthropic is now price-matched with OpenAI at the mid-tier, so choose on quality for your task.

Compared with ChatGPT, Claude trades breadth for depth. You won't get image generation or an ultra-cheap Go tier, but you do get a strong coding agent and writing. Our recommendation: start with the Free tier, test it on real work for a week, and upgrade to Pro if you hit the ceiling. Explore our full AI tools directory to see how Claude fits with the rest of your stack.