Claude can be reached in several different ways, and each way is built for a different kind of person and task.

When most people first hear about Claude, they picture a chat box on a website where you type a question and read an answer. That chat experience is real and it is where most people begin. But it is only the front door. Around it are many other entrances that put Claude exactly where your work already happens, such as inside your code editor, your browser, or your team’s messaging tool.

Anthropic groups these entrances into a few families. There is the core Claude app for everyday use. There are specialized products for engineers, scientists, and security teams. There are add-ons that live inside tools people already use. And there is a developer path, where companies build Claude into their own products. Knowing which family you belong to saves time, because the right entrance feels natural, while the wrong one feels like using a screwdriver to hammer a nail.

flowchart TD
    A["Claude models"] --> B["Core Claude app: web, desktop, mobile"]
    A --> C["Specialist products: Code, Science, Security"]
    A --> D["Inside your tools: Chrome, Microsoft 365, Google Workspace, Slack"]
    A --> E["Developer path: Console and API"]
    A --> F["Cloud platforms: Bedrock, Google Cloud, Microsoft Foundry"]
        

A few simple words come up again and again. Here they are in plain English.

This is the main home of Claude and the best starting point for almost everyone.

The core Claude app is simply called “Claude”. It is where most people spend their time. You can reach it in three ways: on the web through your browser, as a desktop app on your computer, and as a mobile app on your phone or tablet. All three give you the same basic experience of talking with Claude, and your conversations can follow you from one device to another.

Inside this app, Claude does much more than answer questions. It can help you write and edit text, summarize long documents, analyze information, brainstorm ideas, and carry out longer tasks that end in a finished piece of work, such as a report or a document. The app is where you will find the settings that shape your experience, such as which model to use and how much effort Claude should apply.

Within the main app there are features that make Claude more useful for real projects.

The app is more than a plain chat box. Several built-in capabilities help you organize work and produce things you can keep. These are not separate products. They are tools inside Claude, and learning them is like discovering the extra drawers in a desk you already own.

Claude Code is Claude built for the day-to-day work of writing and managing software.

Software engineers spend their days reading code, writing code, fixing bugs, and organizing projects. Claude Code is a product made for exactly this. Instead of copying and pasting snippets into a chat window, an engineer can let Claude work directly with a whole project: reading its files, suggesting and making changes, running checks, and explaining how the pieces fit together.

You can reach Claude Code from the terminal, which is the text-based command window many developers prefer, and also from the web, the desktop app, and mobile. This means an engineer might start a task at their desk and check its progress later from a phone.

Two specialist products serve researchers and security teams with very particular needs.

Claude Science

Claude Science is described as a research workbench for scientists, and it runs on a lab’s own machines. This is an important detail. Scientific work often involves sensitive data, special equipment, and strict rules about where information may travel. Running on the lab’s own computers gives research teams more control over where their work lives while still letting them use Claude to read papers, plan analyses, and organize findings.

Claude Security

Claude Security is aimed at security teams. Its purpose is to find vulnerabilities in a codebase and help patch them. A vulnerability is a weakness in software that an attacker might exploit. Finding and fixing such weaknesses before attackers do is a constant race, and a tool that can help scan code and suggest repairs is valuable to the people defending a system.

Many people do not want a new place to visit. They want Claude to show up in the tools where they already work.

Switching between apps constantly is tiring. Anthropic therefore offers versions of Claude that live inside popular tools. They are named for where they run, which makes them easy to remember.

flowchart LR
    A["You are working in a tool"] --> B{"Which tool?"}
    B -- "Web browser" --> C["Claude in Chrome"]
    B -- "Office documents" --> D["Claude for Microsoft 365"]
    B -- "Docs and sheets" --> E["Claude for Google Workspace"]
    B -- "Team chat" --> F["Claude Tag in Slack"]
    C --> G["Claude helps without leaving the tool"]
    D --> G
    E --> G
    F --> G
        

Connectors let Claude reach information and tools that live in other services you use.

By default, Claude only knows what you tell it and what it learned during training. But much of your real work lives elsewhere: in your calendar, your email, your shared documents, and your project trackers. A connector is a bridge that lets Claude read from, and sometimes act in, those services, but only after you allow it.

For example, instead of copying a list of meetings into the chat, you can connect your calendar and ask Claude about your week. Instead of pasting a document, you can point Claude to it. This saves effort and keeps answers grounded in your real information.

Companies and individual developers can build Claude directly into their own products.

Not everyone wants to use Claude as a finished app. Some people want to build something new with it, such as a customer support assistant, a writing helper inside their own website, or a tool that reads and sorts thousands of documents. For them, Anthropic provides the Claude API and the Claude Console.

The API is the technical connection that lets a program send a request to Claude and receive an answer. The Console is the web dashboard where developers set up their access, try out prompts, and keep track of how much they are using. Together they form the developer path.

Large companies often prefer to reach Claude through the cloud services they already use and trust.

Big organizations frequently have agreements, security rules, and billing arrangements with a major cloud provider. For them, it is simpler to use Claude through that same provider than to set up a new relationship. Claude is therefore available through several large cloud platforms, including Amazon Bedrock, Google Cloud’s Vertex AI, and Microsoft Foundry.

Using Claude this way means the company can keep its AI work inside the cloud environment it already manages. It can use that provider’s tools for access control and monitoring. The trade-off is that some features that depend on Anthropic’s own infrastructure may not be available in every cloud setting, and the data handling is governed by the cloud provider’s rules. Teams should read the specific documentation for their platform before they decide.

Some versions of Claude are tailored to the needs of a whole field of work.

Different professions have their own vocabulary, rules, and habits. Anthropic therefore offers packaged ways of using Claude for particular industries and functions, with names such as Claude for Legal and Claude for Financial Services. Education is another area with its own tailored offering.

These packages are not separate brains. They are the same Claude, arranged and set up with the tools, connections, and guidance that suit a profession. It is like a toolbox that has been pre-packed for a plumber versus one for an electrician. The tools come from the same factory, but the selection fits the job.

If you are not sure where to start, match your situation to the right entrance.

flowchart TD
    A["What do you want to do?"] --> B{"Build software or a product?"}
    B -- "Write code" --> C["Claude Code"]
    B -- "Build Claude into an app" --> D["Claude API and Console, or a cloud platform"]
    B -- "No" --> E{"Work inside another tool?"}
    E -- "Yes" --> F["Chrome, Microsoft 365, Google Workspace, or Slack"]
    E -- "No" --> G["Core Claude app on web, desktop, or mobile"]
        

Whichever door you use, some important things are consistent.

It is reassuring to know that Claude does not become a different assistant in each place. The same family of models sits behind the different surfaces, and Claude keeps the same general character: helpful, curious, careful, and honest. What changes is the setting and the tools around Claude, not its basic nature.

What can differ

Features, layouts, and limits can vary from one surface to another. A feature you love in the main app might not exist inside a particular add-on. Usage allowances and plan details can also differ, and companies can restrict certain options for their staff. When something seems to be missing, a quick look at the official Help Center usually explains why.

A wide choice of places is a strength, but it also asks a little effort from you.

A few sensible habits keep you safe no matter which entrance you use.

Most problems come from picking the wrong entrance or ignoring the rules of the place you chose.

Let us follow an imagined project manager named Meera through one working day to see how the doors fit together.

In a single day Meera used five different doors, each chosen because it was closest to what she was doing. She also kept control the whole time: she checked the numbers, decided what to send, and stayed in charge of her own work.

The list of places where Claude works keeps growing, so it helps to know how to stay current.

New products, new integrations, and new features appear regularly. A tool that was only a preview last year may be widely available now. The best way to keep up is not to memorize a list but to know where to look. The Anthropic Help Center explains how features work in the apps. The developer documentation covers the API and cloud platforms. The release notes and news pages announce what is new.

It is also worth remembering that the ideas in this guide stay steady even when names change. There will always be a main place for everyday use, specialist tools for demanding professions, add-ons that live inside other software, and a developer path for building new products. If you understand those four families, you can make sense of whatever new entrance appears.

flowchart LR
    A["New product or feature appears"] --> B{"Which family does it belong to?"}
    B -- "Everyday use" --> C["Core Claude app"]
    B -- "Demanding profession" --> D["Specialist product"]
    B -- "Inside other software" --> E["Add-on or extension"]
    B -- "Building something new" --> F["Developer path"]
        

Quick answers to the questions beginners ask most often.

The core Claude app on the web or on your phone. It needs no special setup, and everything else builds on the same ideas.

Not for the web version. The desktop and mobile apps are installed like any other app, and browser extensions are added to your browser.

Yes. There is a mobile app, and it includes voice mode so you can talk to Claude. Voice mode is also available for hands-free use through CarPlay.

Claude Code is a product for working with software projects. It is designed for people who write or manage code, so it is most useful to engineers rather than casual chat users.

Yes, through Claude in Chrome, a browser extension that works in your real Chrome browser and can act on pages with your permission.

Yes. There are versions for Microsoft 365 and Google Workspace, and Claude Tag works as a shared assistant in Slack.

The app is a finished product you use directly. The API is a building block that developers use to put Claude inside their own software.

Because it fits the agreements, security controls, and billing they already have with that provider. Amazon Bedrock, Google Cloud, and Microsoft Foundry are examples.

The same family of models and the same general character stand behind the different places. Features, layouts, and limits can differ from one surface to another.

It may depend on the surface you use, your plan, or settings chosen by your organization. Updating your app and checking the Help Center are good first steps.

Connections require your permission, and you can remove them. Connect only services you trust, grant only the access a task needs, and review actions before anything irreversible.

The Anthropic Help Center for the apps, the developer documentation for the API and cloud platforms, and the release notes and news pages for new announcements.

Claude can be reached in many places, and each one is built for a different kind of person and task. The core Claude app on web, desktop, and mobile, with voice mode, is the best starting point for most people. Claude Code, Claude Science, and Claude Security serve engineers, researchers, and security teams. Claude in Chrome, Claude for Microsoft 365, Claude for Google Workspace, and Claude Tag in Slack bring Claude into tools people already use. Developers build with the Claude API and Console, and large organizations can reach Claude through cloud platforms such as Amazon Bedrock, Google Cloud, and Microsoft Foundry. Wherever you meet Claude, the same careful habits apply: choose the right door, grant only the access you need, review important actions, and verify key facts.