Based on Anthropic's own published plan list and data terms, checked in Oct 2026; details may have changed since.
1. What Anthropic charges
Anthropic lists its prices openly, which makes Claude unusual next to most legal AI platforms. The table shows each plan, its published price and the data terms it falls under. Prices are as published in Oct 2026.
| Plan | Published price | Billing unit | What matters to a lawyer |
|---|---|---|---|
| Free | $0 | Per user | Consumer terms: training use depends on a privacy setting |
| Pro | $20 per month, or $17 billed annually | Per user | Consumer terms: training use depends on a privacy setting |
| Max | from $100 per month (5x and 20x usage tiers) | Per user | Consumer terms: training use depends on a privacy setting |
| Team, standard seat | $25 per seat per month (standard seat), or $20 billed annually | Per seat, minimum 2 seats | Commercial terms: no training on customer content |
| Team, premium seat | $125 per seat per month, or $100 billed annually | Per seat | Commercial terms: no training on customer content |
| Enterprise | $20 per seat per month billed annually, plus usage at API rates | Per seat plus usage | Commercial terms, plus audit logs, role-based access and custom data retention controls |
Higher plans mainly raise how much a person can use Claude before reaching a limit. Anthropic also sells per-token API access for developers, which is a separate price list and not how most firms buy it.
2. What each plan means for client data
For a law firm the more important difference between the plans is the data terms, not the price. Anthropic runs two sets of terms.
Consumer plans (Free, Pro, Max).Anthropic uses chats to improve Claude only if the user allows it in a privacy setting, if a conversation is flagged for safety review, or if the user opts in in another way, such as a testing program. Where a user allows it, Anthropic may keep the data in de-identified form for up to five years. Otherwise, deleted chats leave Anthropic's back-end storage within 30 days. The setting belongs to each individual account, so a firm cannot assume that every lawyer has it switched the same way.
Commercial plans (Team, Enterprise, API).Under Anthropic's commercial terms, Anthropic may not train its models on customer content from these services. Deleted chats and API inputs and outputs are removed from back-end storage within 30 days, with exceptions for usage policy violations and legal requirements. Enterprise adds audit logs, role-based access and custom data retention controls, and Anthropic lists a HIPAA-ready option on that plan.
None of this replaces the firm's own duty of confidentiality. Read the current terms and your jurisdiction's ethics guidance before client material goes into any AI tool.
3. What Claude for legal is, and is not
Searches for “Claude for legal” and “Claude for lawyers” describe a law firm using Claude's general plans for legal work. None of the plans in the table is priced or packaged as a legal product, and Anthropic's product pages do not document a built-in legal citator as of this review.
That does not make Claude unfit for legal work. Lawyers use it to draft, summarize and reason through documents they supply. It does mean the firm supplies the legal workflow: the research sources, the checking, the matter structure and the review. The price on the list covers the model access, not those parts.
4. What a firm adds on top
Citation checking. Every authority Claude cites has to be confirmed as real and still good law before it is relied on. With Claude alone that is a manual step or a separate tool, and it is a cost in attorney time, not in the plan price.
Matter memory. A matter runs for months across many documents and many sessions. A firm using Claude has to decide where that context lives and how it is carried into each new conversation. Some legal platforms build that into the product; with a general assistant the firm builds it.
For a side-by-side view, see Harvey vs Claude and Legora vs Claude. For the wider decision across platforms, see the real decision between Harvey, Legora, Claude and Irys. For how other legal AI platforms price, see the legal AI pricing comparison.
5. Questions to settle before you budget
- Which plan will every lawyer who touches client material be on, and is that plan under commercial terms?
- Is anyone on a personal Free, Pro or Max account for client work, and what is their model improvement setting?
- How will the firm check every citation Claude produces before it reaches a filing or a client?
- Where will matter context live between sessions, and who can see it?
- What happens when a heavy workload reaches the usage limit of a plan?
- Does the firm's engagement letter or ethics guidance require client consent or an outside-counsel policy for AI tools?
Frequently asked questions
How much does Claude cost for a lawyer?
Anthropic publishes the prices. A solo lawyer on a personal plan pays $0 on Free, $20 per month, or $17 billed annually on Pro, or from $100 per month on Max. A firm on Team pays $25 per seat per month (standard seat), or $20 billed annually, or $125 per seat per month, or $100 billed annually for a premium seat. Enterprise is $20 per seat per month billed annually, plus usage at API rates.
Is there a Claude plan made for lawyers?
Not on the pricing list Anthropic publishes. The plans are general plans. What makes the work legal is the workflow a firm builds around them: which documents go in, how citations are checked, and how a matter's context carries between sessions.
Does Anthropic train Claude on a law firm's data?
It depends on the plan. Under Anthropic's commercial terms, which cover Team, Enterprise and the API, Anthropic may not train models on customer content. On the consumer plans (Free, Pro, Max), chats are used to improve Claude only if the user allows it in a privacy setting, or in the other cases Anthropic lists, such as a conversation flagged for safety review.
How long does Anthropic keep a firm's chats?
Anthropic states that deleted chats leave its back-end storage within 30 days, and that API inputs and outputs are deleted within 30 days, with exceptions for policy violations and legal requirements. On consumer plans where the user allows model improvement, data may be kept in de-identified form for up to 5 years. Enterprise plans list custom data retention controls.
Does Claude check legal citations?
Anthropic's product pages do not document a built-in legal citator as of this review. A firm using Claude needs its own process, or a separate tool, to confirm that each cited authority exists and is still good law.
Is Claude cheaper than a legal AI platform?
Claude's published Team price is $25 per seat per month (standard seat), or $20 billed annually, against the $299 per month per seat that Irys charges. The comparison is incomplete without the work a firm adds: citation checking, matter context and review. Harvey and CoCounsel do not publish a standard price, so there is no published figure to compare against; request quotes for your firm size and compare the total.
Who publishes this guide?
Irys (formerly Iqidis), a legal AI platform with published pricing: $299 per month per seat, with a 7-day free trial. Irys competes with general-purpose assistants like Claude at the smaller-firm end of the market; this guide keeps its claims to what Anthropic's own pages state.
This guide is based on Anthropic's own published plan list and data terms, reviewed in Oct 2026. Prices, plan limits and terms change frequently; confirm current details with Anthropic before making a purchasing decision. Claude and Anthropic are trademarks of their owner, used here for identification and comparison only, with no affiliation or endorsement implied. Irys (formerly Iqidis) publishes this guide and competes with general-purpose assistants for legal work. This guide is informational and is not legal advice.
