Quick answer
Claude is a large language model (LLM) built by Anthropic. As of October 7, 2026, Anthropic's help center lists writing, summarizing, translating and coding among its uses and says Claude can occasionally produce incorrect or misleading responses; its pricing page lists the Free plan at "$0".
Key takeaways
- Claude is one company's product: the help center calls it a large language model built by Anthropic, and the Free plan is listed at $0 as of October 7, 2026.
- Anthropic's help center lists writing, learning, summarizing, coding, brainstorming, translating and image analysis as uses, and says Claude can occasionally produce responses that are incorrect or misleading.
- For Free, Pro and Max, Anthropic's privacy article lists "You choose to allow us to use your chats and coding sessions to improve Claude" as a condition for training, while its pricing page labels model training "Opt-out"; check the toggle on your own account. Incognito chats are not used to improve Claude, even if you have enabled Model Improvement.
- This site's suggestion: try the Free plan on text that is not confidential, check the privacy toggle first, and verify any fact before you rely on it.
What is Claude AI?
Claude is a large language model (LLM) built by Anthropic, and Anthropic's help center lists writing, summarizing, translating and coding among the things you can use it for.
Disclosure: this page was drafted with Claude, the product it describes, made by Anthropic.
In this site's plain words, not Anthropic's: you type a request in everyday language and it writes a reply. Claude is one company's product, not a name for AI in general; the definition and uses come from Anthropic's help center, read on October 7, 2026.
Where would a business owner meet Claude?
On the web, in desktop and phone apps, and, per Anthropic's product page, inside tools you connect to it (quoted under the uses below). The help center says: "Web : Visit claude.ai to chat with Claude in a web browser. Desktop : Download Claude Desktop for Mac or Windows. Mobile : Download Claude for iOS or Android to chat on your mobile device."
Two conditions apply. The same page says "You must be in one of our supported locations to access Claude." and "You must be at least 18 years old to use our services."
As of October 7, 2026, Anthropic's pricing page lists the Free plan at "$0"; other plans and prices are on that page.
What does Anthropic say Claude is for?
Anthropic's help center lists writing, learning, summarizing, coding, brainstorming, translating and analyzing images. Its own labels are "Write across different formats and styles.", "Summarize text." and "Analyze images.", alongside coding, brainstorming and learning something new.
Its product page says: "Claude takes real actions on your files and in connected tools. You choose which connectors Claude can use — it works only in the folders and tools you allow." It adds that Claude "can produce real files like documents, spreadsheets, and slides that you review, edit, and sign off on." As of October 7, 2026, the pricing page's Free plan card lists, among other items, "Search the web, create files, and run code", "Memory across conversations" and "Connect your apps and tools".
On translation the help center says: "While Claude is best at English due to its training data, Claude knows more than a dozen languages and can translate between them to varying degrees of success."
For one task in practice, see AI meeting notes for small business; for the wider picture, the small business AI automation guide.
What does Anthropic say about Claude's mistakes and limits?
It says Claude can be wrong, that it may not know recent events, and that usage is capped. Anthropic's help center: "Claude can occasionally produce responses that are incorrect or misleading." It adds: "In other words, Claude can write things that might look correct but are very mistaken." Its advice: "Users should not rely on Claude as a singular source of truth and should carefully scrutinize any high-stakes advice given by Claude."
When Claude searches the web, the same page says "users should review Claude's cited sources." On recent events, another help page says "each model has a knowledge cutoff", "These models may not be aware of events or information that occurred after their respective cutoff dates." and "If you ask about more recent events, the model may not have accurate information."
On usage, the help center says: "While using the free Claude plan, there is a session-based usage limit that will reset every five hours." The page gives no message count.
What happens to my chats?
On Free, Pro and Max, Anthropic's pricing page labels training "Opt-out" and its privacy article lists "You choose to allow us to use your chats and coding sessions to improve Claude" as a condition; check your account's toggle.
The privacy article, dated March 16, 2026 on the page, covers "consumer products such as Claude Free, Pro, Max". It lists three conditions for training: that choice; "Your conversations are flagged for safety review" (in its words, "we may use or analyze them to improve our ability to detect and enforce our Usage Policy, including training models for use by our Safeguards team"); and "By otherwise explicitly opting in to training".
The retention article words its flag differently (see the retention table below), and the pages do not say whether these are the same flag. Neither states your account's toggle setting, so look at the toggle itself ("Help Improve our AI models").
| Question | What Anthropic's page says |
|---|---|
| If I turn it off? | "we will not use any new chats and coding sessions you have with Claude for future model training"; "Your data will still be included in model training that has already started and in models that have already been trained, but we will stop using your previously stored chats and coding sessions in future model training runs." |
| If I turn it off? | "we will not use your previous or new chats or coding sessions for future model training" |
| If I delete a chat? | "if you delete a chat from your chat history, it will not be used to train future models" |
| Flagged chats, with it off? | "If our safety classifiers flag your conversations, they may still be used to improve our internal trust and safety models, detect harmful content, enforce our policies, or advance our safety research." |
The periods the retention and training articles state:
| Situation | Period as the page states it | What it applies to |
|---|---|---|
| Deleted conversation | "Removed from your chat history immediately"; "Deleted from our back-end storage systems within 30 days" | A conversation you delete |
| Training allowed | "up to 5 years" in a de-identified format, in "model training pipelines" | "This retention period only applies to new or resumed chats after the model training setting has been enabled." |
| Flagged for Usage Policy violation | Inputs and outputs "up to 2 years"; trust and safety classification scores "up to 7 years" | A chat or session "flagged by our automated trust and safety systems as violating our Usage Policy" |
| Feedback | Thumbs up/down: "up to 5 years" (privacy article). Thumbs up/down "or sent bug reports": "for 5 years" (retention article) | Privacy article: "the entire related conversation" for thumbs up/down. Retention article: "data associated with that submission" |
| Incognito chat | "30 days (default)"; "Your Incognito chats are not used to improve Claude" | "temporary conversations that aren't saved to your chat history or to Claude’s memory" |
The retention article adds: "In all cases, we may retain chats and coding sessions as required by law, to resolve disputes, or as necessary to combat violations of our Usage Policy."
What is a sensible first try, and what should I check?
This site's suggestion, not Anthropic's: start on the Free plan with text that is not confidential, and run these steps in order.
- Check the two conditions: 18 or over, and a supported location.
- On Free, Pro or Max, before pasting anything about customers, open the privacy toggle and see how it is set, or use an incognito chat.
- Expect the Free plan's five-hour session limit. The help center gives no message count; it says paid plans offer higher limits.
- Give it one real task from Anthropic's own list, such as summarizing a long email thread or drafting a reply.
- Check the output against the source: confirm names, figures and dates yourself.
FAQ
Who makes Claude AI?
Anthropic. The Claude help center says, "Claude is a large language model (LLM) built by Anthropic."
Is Claude AI free to use?
There is a free plan. As of October 7, 2026, Anthropic's pricing page lists Free at "$0" and says "Free for everyone". The help center adds that a session-based usage limit applies on it.
What is Claude AI used for?
Anthropic's help center lists writing, learning something new, summarizing text, coding, brainstorming, translating languages and analyzing images.
Does Claude use my chats to train its models?
For Free, Pro and Max, Anthropic's privacy article lists "You choose to allow us to use your chats and coding sessions to improve Claude" among its conditions, and its pricing page labels model training "Opt-out"; check the toggle on your own account. Flagged conversations are covered separately in the privacy article, quoted on this page. Incognito chats are not used to improve Claude, even if you have enabled Model Improvement. The pricing page lists Team with no model training on your content by default.
Is Claude better than ChatGPT?
These pages give no basis to rank them: they report only Anthropic's own statements and ran no test.
How this page was made
This page was drafted with Claude, an AI model made by Anthropic, and checked against each vendor's published material on October 7, 2026. Page dates are as each page prints them: privacy article March 16, 2026; settings article August 3, 2026; retention article July 1, 2026. Every statement about Claude comes from Anthropic's pricing, product, help-center and privacy-center pages, opened that day and read rendered in a browser from outside the United States; no third-party review or benchmark was used. We ran no quality test. Plans, prices and wording change often, so check the linked pages.