Claude is a family of AI assistants built by Anthropic. In its most familiar form, Claude is a conversational assistant you use at claude.ai and through desktop and mobile apps: you describe a task in plain language and Claude responds. It is strong at writing, editing, summarizing, answering questions, and analyzing content you give it. The name Claude also refers to the underlying models, which power Anthropic's other Claude products. So Claude is both the chat assistant people talk to and the engine behind a wider product lineup.

Claude is made by Anthropic, an AI safety and research company. Anthropic builds the Claude models and the products that run on them, including the claude.ai chat assistant, the desktop and mobile apps, a delegated-work assistant, a coding agent, and the Claude API that lets developers build their own applications on the same models. Lightbridge Labs is an independent practice and is not affiliated with, endorsed by, or a partner of Anthropic.

What are the Claude models?

Claude comes in tiers that trade capability against speed. The most capable tier is for hard reasoning and deep analysis; in the mid-2026 lineup it includes Claude Fable 5 as the most capable model alongside Claude Opus 4.8. The balanced tier, with Claude Sonnet 4.6, is the default for most everyday work. The fast tier, with Claude Haiku 4.5, suits high-volume, simpler tasks where speed matters. Because the lineup changes quickly, confirm the current models and their capabilities at docs.anthropic.com and anthropic.com before planning around a specific one.

Claude offers a free tier at claude.ai with limited usage, plus paid plans that unlock more usage, access to the more capable models, longer context, and team and enterprise administration features. The exact plan structure and what each tier includes change over time, so check anthropic.com for the current options. Lightbridge Labs does not resell Claude plans; this guide is independent, general information.

What is Claude good at?

As a conversational assistant, Claude is strong at writing and editing, summarizing documents, explaining concepts, answering questions, drafting communications, and analyzing content you paste into the conversation. It is well suited to one-off tasks that do not need to connect to your own files or systems. For work that spans a folder of files or a whole code project, Anthropic offers other Claude products built for those jobs, which Lightbridge Labs covers in its Claude products compared guide at https://lightbridgelabs.ai/resources/claude-products-compared.

What are Claude's limits?

The claude.ai chat assistant only sees what you put into the conversation: it cannot reach your files, run code, or act on external systems on its own. By default it does not carry memory across separate conversations, so each one starts fresh. Like any AI assistant it can be confidently wrong, so important outputs deserve a human check. For tasks that need direct file access or autonomous multi-step work, Anthropic's other Claude products fit better. Getting better results from any of them comes down to clear prompting, covered in the Lightbridge Labs guide at https://lightbridgelabs.ai/resources/how-to-prompt-claude.

How is Claude different from other AI assistants?

Claude is Anthropic's assistant, and Anthropic positions itself as an AI safety and research company, so a focus on helpful, honest, and careful responses is central to how Claude is designed. Beyond the chat assistant, the Claude family extends into a delegated-work assistant for knowledge workers and a coding agent for developers, all running on the same underlying models, with a developer API underneath. The practical way to compare assistants is against your own tasks rather than a feature list. Lightbridge Labs helps organizations make that call through its AI strategy work at https://lightbridgelabs.ai/consulting/ai-strategy.

RL Written by Robert LabardeeFounder and CEO

Last updated June 2026.

What is Claude?

Lightbridge Labs defines Claude as the family of AI assistants built by Anthropic. At its core Claude is a conversational assistant you reach at claude.ai and through desktop and mobile apps. The same models also power Anthropic's other Claude products, so Claude names both the chat assistant and the engine behind the wider lineup.

Compare the Claude products Learn how to prompt Claude

Claude is the family of AI assistants built by Anthropic.

Claude is an AI assistant created by Anthropic, an AI safety and research company. In its most familiar form, Claude is the conversational assistant at claude.ai: you describe a task in plain language and it responds with writing, analysis, or answers. The word Claude carries two meanings at once. It is the chat assistant most people talk to, and it is also the name of the underlying models that power Anthropic's other Claude products.

That distinction matters when choosing where to work. The chat assistant is built for conversation and one-off tasks. The same models, presented through other products, handle delegated work across a folder of files or hands-on coding in a project. This guide stays definitional and broad. For a side-by-side of which product fits which job, see the Lightbridge Labs Claude products compared guide and the rest of the resources hub.

Claude models come in tiers that trade capability against speed.

Anthropic offers Claude in a small set of model tiers rather than a single model. The pattern is consistent even as the names change: a most-capable tier for the hardest work, a balanced tier that handles most tasks, and a fast tier for high-volume, simpler jobs. Picking a tier is a trade between depth of reasoning and how quickly an answer comes back.

Most capable tier

The strongest models, built for hard reasoning, deep analysis, and the most demanding work. In the mid-2026 lineup this tier includes Claude Fable 5, the most capable model, alongside Claude Opus 4.8. Reach for this tier when accuracy on a complex problem matters more than speed.

Balanced tier

The default for most everyday work, pairing strong quality with quick responses. Claude Sonnet 4.6 sits here in the mid-2026 lineup. It handles the majority of writing, analysis, and question-answering tasks without reaching for the heaviest model.

Fast tier

The quickest and lightest models, suited to high-volume, simpler tasks such as classification, extraction, and short responses where latency matters. Claude Haiku 4.5 is the mid-2026 fast tier model.

Model names and capabilities move quickly. Treat the names above as a snapshot of the mid-2026 lineup, and confirm the current models, context, and limits at docs.anthropic.com and anthropic.com before planning around a specific one.

Claude runs at claude.ai, in desktop and mobile apps, and behind the wider product family.

Most people meet Claude through the chat assistant, but the same models reach further. Claude is available in the browser, as desktop and mobile apps, and as the engine inside Anthropic's other Claude products and the developer API. These are the main places Claude shows up.

claude.ai chat

The conversational interface at claude.ai and what most people mean by Claude. You type, Claude responds. There is nothing to install, and it is the fastest way to start.

Desktop and mobile apps

Claude runs as a desktop application and on iOS and Android, so the same assistant is available on a laptop or a phone. The apps also host Anthropic's other Claude products.

Engine behind the family

The same Claude models power Anthropic's other products, including the delegated-work assistant and the coding agent, and they are available programmatically through the Claude API.

Claude is strong at language work, and its limits are worth knowing.

As a conversational assistant, Claude is at its best with language. It writes and edits, summarizes long documents, explains concepts, answers questions, drafts communications, and analyzes content you paste into the conversation. It is well suited to one-off tasks that do not need to touch your own files or systems, and getting strong results comes down mostly to clear prompting, which the Lightbridge Labs how to prompt Claude guide covers.

The limits follow from the same design. The chat assistant only sees what you put into the conversation: it cannot reach your files, run code, or act on outside systems by itself, and by default it does not carry memory from one conversation to the next. Like any AI assistant, it can be confidently wrong, so important outputs deserve a human review. When a task needs direct file access or autonomous, multi-step work, a different member of the family fits better.

Claude the assistant is the entry point to a wider Claude product family.

The claude.ai assistant is one part of a lineup that all runs on the same Claude models. Alongside it, Anthropic offers a delegated-work assistant for knowledge workers who want a goal completed across a folder of files, and a coding agent built to read, write, and run code in a project. Underneath them all sits the Claude API, which lets developers build their own applications on the same models.

Choosing among them is a fit question, not a ranking. The Lightbridge Labs Claude products compared guide sets each product against the others, what is Claude Cowork explains the delegated-work assistant, and what is Claude Code covers the coding agent. For help putting Claude to work across an organization, Lightbridge Labs offers AI strategy guidance.

What is Claude: frequently asked questions

What is Claude?: Claude is a family of AI assistants built by Anthropic. In its most familiar form, Claude is a conversational assistant you use at claude.ai and through desktop and mobile apps: you describe a task in plain language and Claude responds. It is strong at writing, editing, summarizing, answering questions, and analyzing content you give it. The name Claude also refers to the underlying models, which power Anthropic's other Claude products. So Claude is both the chat assistant people talk to and the engine behind a wider product lineup.
Who makes Claude?: Claude is made by Anthropic, an AI safety and research company. Anthropic builds the Claude models and the products that run on them, including the claude.ai chat assistant, the desktop and mobile apps, a delegated-work assistant, a coding agent, and the Claude API that lets developers build their own applications on the same models. Lightbridge Labs is an independent practice and is not affiliated with, endorsed by, or a partner of Anthropic.
What are the Claude models?: Claude comes in tiers that trade capability against speed. The most capable tier is for hard reasoning and deep analysis; in the mid-2026 lineup it includes Claude Fable 5 as the most capable model alongside Claude Opus 4.8. The balanced tier, with Claude Sonnet 4.6, is the default for most everyday work. The fast tier, with Claude Haiku 4.5, suits high-volume, simpler tasks where speed matters. Because the lineup changes quickly, confirm the current models and their capabilities at docs.anthropic.com and anthropic.com before planning around a specific one.
Is Claude free?: Claude offers a free tier at claude.ai with limited usage, plus paid plans that unlock more usage, access to the more capable models, longer context, and team and enterprise administration features. The exact plan structure and what each tier includes change over time, so check anthropic.com for the current options. Lightbridge Labs does not resell Claude plans; this guide is independent, general information.
What is Claude good at?: As a conversational assistant, Claude is strong at writing and editing, summarizing documents, explaining concepts, answering questions, drafting communications, and analyzing content you paste into the conversation. It is well suited to one-off tasks that do not need to connect to your own files or systems. For work that spans a folder of files or a whole code project, Anthropic offers other Claude products built for those jobs, which Lightbridge Labs covers in its Claude products compared guide at https://lightbridgelabs.ai/resources/claude-products-compared.
What are Claude's limits?: The claude.ai chat assistant only sees what you put into the conversation: it cannot reach your files, run code, or act on external systems on its own. By default it does not carry memory across separate conversations, so each one starts fresh. Like any AI assistant it can be confidently wrong, so important outputs deserve a human check. For tasks that need direct file access or autonomous multi-step work, Anthropic's other Claude products fit better. Getting better results from any of them comes down to clear prompting, covered in the Lightbridge Labs guide at https://lightbridgelabs.ai/resources/how-to-prompt-claude.
How is Claude different from other AI assistants?: Claude is Anthropic's assistant, and Anthropic positions itself as an AI safety and research company, so a focus on helpful, honest, and careful responses is central to how Claude is designed. Beyond the chat assistant, the Claude family extends into a delegated-work assistant for knowledge workers and a coding agent for developers, all running on the same underlying models, with a developer API underneath. The practical way to compare assistants is against your own tasks rather than a feature list. Lightbridge Labs helps organizations make that call through its AI strategy work at https://lightbridgelabs.ai/consulting/ai-strategy.

This guide is independent, general educational information published by Lightbridge Labs. Claude and Anthropic are trademarks of Anthropic, PBC. Lightbridge Labs is not affiliated with, endorsed by, or a partner of Anthropic, PBC.

From understanding Claude to putting it to work.

When the question shifts from what Claude is to how your organization should adopt it, Lightbridge Labs helps you choose the right products, set the strategy, and build the custom capabilities around them.

Talk to Lightbridge Labs Compare the Claude products