What you will be able to do
- Decide whether a job calls for plain chat with web search, Research, a Project, an Artifact or Cowork
- State the plan requirements and prerequisites for Research, Projects, RAG, Artifacts and Cowork
- Explain what a Project shares across its chats and what it does not
- Recognise when Claude will produce an artifact and how artifacts are edited, saved and shared
Key concept
Match the feature to the shape of the job — Each Claude feature fits a particular kind of work. Use a plain chat for a quick answer, Research for synthesis across many sources, a Project for context you reuse, and an Artifact for a deliverable you will keep editing. To choose, ask what the work looks like, not which feature is most powerful.
1.Start from chat: the quick-answer baseline
Every other feature in this lesson builds on a plain conversation with Claude. You ask, Claude answers, and the exchange stays inside that one chat. Most questions need nothing more. When the answer depends on current information, you can turn on web search in the same chat. Anthropic's guidance describes web search as the right tool for straightforward, factual queries that one or two tool calls can answer. Examples include the weather, details about one company, recent headlines, or last night's game result.
Web search also suits you when you want to steer. You tell Claude roughly what to look up, and Claude runs the searches and analyses the results. Extended thinking covers the opposite case: hard reasoning, such as maths, debugging or weighing ethical arguments, where nothing needs to come from the web. Extended thinking on its own is the choice when you only want Claude to think harder about the question, with no searching and no integrations.
So the baseline already gives you three settings: a plain answer, a quick lookup with web search, or deeper reasoning with extended thinking. The features in the rest of the lesson cover jobs this baseline handles poorly. Those are investigations that need many searches, context you would otherwise re-upload every time, and output you will keep refining and hand to someone else.
Checkpoint 1 of 8· Check yourself
A user wants to know who won a particular sports match last night and is happy to tell Claude roughly where to look. Which capability fits best?
One or two tool calls can answer a single current fact, and that is the job web search is for. Research is meant for multi-source synthesis, and extended thinking cannot fetch recent information.
“Web search is best for straightforward, factual queries that can be answered with one or two tool calls”Source: support.claude.com
Sources1
2.Research: agentic, multi-source investigation with citations
Research is what you use when one or two searches will not settle the question. Claude works agentically. It runs several searches, and each one builds on the last. Claude decides what to investigate next, explores different angles of your question, and works through open questions systematically. Anthropic positions it for gathering that takes five or more tool calls over one to three minutes. The output is a longer report that brings many sources together with proper citations, so you can check each claim against where it came from.
Research is not limited to the public web. When you have connected internal sources, such as Gmail, Google Calendar and Google Docs, Claude searches them alongside the web. That makes it suitable for jobs like comparing business competitors, updating out-of-date internal documents with current information from the web, or prioritising action items from your calendar and email. You can also combine Research with extended thinking. Claude then plans its approach carefully before it gathers information, which suits work like analysing several scientific papers.
| Aspect | What the documentation says |
|---|---|
| Plans | Pro, Max, Team or Enterprise (paid plans only) |
| Surfaces | Claude on the web, Claude Desktop, Claude Mobile |
| Prerequisite | Web search must be turned on |
| Sources searched | The web, plus internal context such as Gmail, Google Calendar and Google Docs when connected |
| Usage | Same limits as standard conversations, but can use them up faster |
To turn it on, click the plus button at the bottom left of the chat box and choose Research. A blue indicator appears at the bottom of the chat window, and clicking it again turns Research off. Then ask your question. Research has a cost. It counts against the same limits as an ordinary conversation, but it uses them up faster because Claude retrieves many sources and writes a full report. Don't use it for a question that web search answers in one call.
Checkpoint 2 of 8· Put it in order
Put the steps for running a Research query in order.
- 1.Confirm the blue indicator appears at the bottom of the chat window
- 2.Click the plus button at the bottom left of the chat interface and choose Research
- 3.Ask Claude your question
- 4.Make sure web search is turned on
Research only works when web search is on. Next you turn Research on from the plus menu, the blue indicator confirms it is active, and then you ask your question.
“You must have web search turned on for research to function.”Source: support.claude.com
Checkpoint 3 of 8· Exam question
A legal team repeatedly asks Claude to draft memos that reference the same set of contract templates and a shared style guide. They want every new conversation to automatically have access to those documents without re-uploading them each time. Which Claude product feature best fits this need?
Correct answer: A — Create a Project and upload the contract templates and style guide to its project knowledge
- A. Correct. Projects maintain a persistent knowledge base that Claude references across all conversations within that project, so uploading the templates and style guide once makes them available to every future memo without re-uploading.
- B. Pasting the same documents into every new conversation defeats the purpose of persistent context and requires manual repetition each time.
- C. Research mode performs agentic web and connected-source searches; it does not give Claude access to private documents that were never uploaded or connected.
- D. An artifact is a standalone content window for a single conversation's output, not a mechanism for supplying persistent background knowledge across future conversations.
3.Projects: shared context for recurring work
Research is for one large investigation. Projects are for the opposite situation: work you come back to repeatedly. A Project is a self-contained workspace with its own chat history and knowledge base. You upload documents, text, code or other files to the Project's knowledge base once, and Claude uses them as background in every chat inside that Project. You can also set project instructions that shape every chat in the Project, for example a formal tone or answering from the point of view of a particular role or industry.
Note exactly what is shared. The knowledge base and the instructions apply to every chat in the Project. The chats themselves do not share context with each other. If you paste a document into one chat, a different chat in the same Project will not see it. Information carries across chats only when you add it to the project knowledge base. The name and description you give a Project are labels for you, and Claude cannot see them.
Projects are available on every plan, including Free. Free users can create up to five. Paid plans add retrieval augmented generation (RAG). When project knowledge approaches the context window limit, Claude switches to RAG mode automatically and searches your files for the relevant parts instead of loading everything at once. This can expand capacity by up to 10x, with no setup. Team and Enterprise plans add sharing. A creator can share a Project with specific people, by bulk email list, or with the whole organisation, using two permission levels: Can view and Can edit.
| Capability | Free | Pro / Max | Team / Enterprise |
|---|---|---|---|
| Create projects with knowledge base and instructions | Yes, maximum of five projects | Yes | Yes |
| RAG mode for large project knowledge | No | Yes | Yes |
| Share projects with organization members (Can view / Can edit) | No | No | Yes, unless an admin has turned sharing off |
Consider a team where everyone keeps re-uploading the same reference files into their own chats. The fix is to put those files in one Project's knowledge base and share the Project. Sharing requires a Team or Enterprise plan, and members with Can view access can still chat inside the Project.
Checkpoint 4 of 8· Check yourself
Inside a Project, a user uploads a policy PDF into Chat A. Later, in Chat B of the same Project, Claude doesn't seem to know the policy. What explains this, and what is the fix?
A file uploaded inside one chat stays in that chat. Only files in the project knowledge base are available to every chat in the Project.
“Context is not shared across chats within a project unless the information is added into the project knowledge base.”Source: support.claude.com
Checkpoint 5 of 8· Exam question
A developer asks Claude to build a small interactive React component that visualizes sales data with adjustable filters, and expects to iterate on the code across several follow-up messages. Which feature will display this output in a way that supports iterative editing and version history?
Correct answer: A — The artifact panel, which shows the component separately and tracks versions as it is revised
- A. Correct. Artifacts open substantial, self-contained content like a React component in a dedicated side panel and provide a version selector, letting the developer iterate and switch between revisions.
- B. Project knowledge is meant for background reference documents the user uploads, not for displaying and iteratively editing generated code with version tracking.
- C. Research mode is for gathering external information through web searches, not for rendering or iterating on generated code.
- D. Printing code inline in the main transcript does not provide the dedicated side-panel view or version history that iterative component development benefits from.
4.Artifacts: durable, editable deliverables
Projects keep your inputs in one place. Artifacts do the same for your outputs. An artifact is something Claude makes that you would show to someone else: a design, a deck, a document, a dashboard or a small interactive tool. It opens beside the conversation, so you can keep chatting while it stays visible. You can edit it, return to it later and share it. Claude creates an artifact when the content is substantial and self-contained (usually more than 15 lines), likely to be edited, iterated on or reused outside the conversation, understandable without the chat around it, and likely to be referred back to.
You can change an artifact in several ways. You can ask Claude to change it. In artifacts made from a template, you can type, edit a slide or move elements on a canvas yourself, and the artifact updates live. In a document, you can highlight a passage and choose Edit with Claude, so Claude knows exactly which part you mean. If you want to try a different direction without losing your current work, edit an earlier message. That creates a separate version of the chat with its own artifacts. Everything you make is saved in the Artifacts tab in the sidebar, so you can reach it from any conversation.
| Feature | Free | Pro | Max | Team | Enterprise |
|---|---|---|---|---|---|
| Create artifacts in a chat | Yes | Yes | Yes | Yes | Yes |
| Start from a template (Design, Slides, Docs) | No | Yes | Yes | Yes | Yes (off until an owner turns each on) |
| Connect your apps to an artifact | No | Yes | Yes | Yes | Yes |
| Store data in an artifact | No | Yes | Yes | Yes | Yes |
Two practical details matter. First, artifacts need Cloud code execution and file creation turned on under Settings > Capabilities, or under Organization settings > Capabilities on Team and Enterprise. Second, what you can do depends on the surface. Web and Desktop let you create, edit and share artifacts and start from templates. On iOS and Android you can ask for an artifact and view it, but templates, editing and sharing settings require web or Desktop. Artifacts can also call Claude directly and act as small apps. When other people use a shared app, the usage counts against their own plan limits, not yours.
Checkpoint 6 of 8· Check yourself
A Free-plan user asks Claude for a small interactive tool, but no artifact appears beside the conversation. What should they check first?
Free plans can create artifacts in a chat. Only templates and some extras are paid. Artifacts do require the code execution and file creation capability to be switched on.
“Artifacts require Cloud code execution and file creation to be turned on in Settings > Capabilities”Source: support.claude.com
Checkpoint 7 of 8· Exam question
A support team lead wants several colleagues to be able to view and use the same set of onboarding documents and a shared system of custom instructions when chatting with Claude, without each person needing to configure their own settings. Which feature addresses this on a Team or Enterprise plan?
Correct answer: A — A shared Project that team members are added to with view or edit permissions
- A. Correct. Team and Enterprise projects can be shared with specific members or the whole organization, with permission levels like 'Can use' and 'Can edit', giving everyone consistent access to the same knowledge base and instructions.
- B. Copying and distributing a single chat transcript does not give colleagues an ongoing shared, editable space with consistent instructions and knowledge.
- C. Research mode searches the web and connected sources on demand; it is not a mechanism for sharing a persistent internal knowledge base and instructions among team members.
- D. Exporting and emailing an artifact distributes a static file rather than providing an ongoing shared workspace with consistent instructions.
Sources6
5.Putting it together: picking the feature for the job
There is one more option. Claude Cowork brings Claude Code's agentic capabilities to knowledge work beyond coding, and it is available on the paid plans: Pro, Max, Team and Enterprise. The documentation also notes that Cowork and chat are merging on some Pro and Max accounts. On those accounts you ask for what you need, and Claude decides whether to give a quick answer or take it on as a task. For this objective, think of Cowork as the choice when you want Claude to carry out a task, not just answer a question.
In practice, look at what the request involves. A single fact or a quick lookup is a chat, with web search if it needs current information. A question that needs many searches, connected email or calendar data, and a cited report is a job for Research. Context you would otherwise upload again and again belongs in a Project. A deliverable you will keep refining and then share should be an Artifact. These features can be combined. Research works inside a RAG-enabled Project, and any chat can produce an artifact.
| Feature | Best for | Plan | Prerequisite or limit |
|---|---|---|---|
| Chat + web search | Straightforward factual queries answered in one or two tool calls | All plans | Web search turned on |
| Research | Multi-source reports with citations, five or more tool calls over 1-3 minutes | Pro, Max, Team, Enterprise | Web search turned on; uses limits faster |
| Projects | Recurring work with a shared knowledge base and instructions | All plans (Free: max five) | Context shared only via project knowledge base |
| Artifacts | Self-contained content you will edit, reuse and share | All plans (templates paid only) | Cloud code execution and file creation turned on |
| Cowork | Agentic knowledge work beyond coding | Pro, Max, Team, Enterprise | Availability varies by surface |
Checkpoint 8 of 8· Match them up
Match each need to the feature that fits it.
Tap a term, then the definition that fits it.
Each need maps to a different kind of work: multi-source synthesis, reused context, a durable deliverable, or a quick factual lookup.
“It can do things like compare business competitors, update outdated internal documents with info from the web”Source: support.claude.com
Exam traps
Each one states something that sounds right. Open it to see what is actually true.
1.Turning on web search is enough to use Research on any plan.Why is that wrong?
Research has two requirements: a paid plan (Pro, Max, Team or Enterprise) and web search turned on. Free users cannot use it.
Covered in Research: agentic, multi-source investigation with citations
2.Projects are a paid feature, so a Free user cannot organise related chats into a dedicated workspace.Why is that wrong?
Projects are available on every plan, including Free, which allows up to five. Only RAG expansion is paid, and sharing requires Team or Enterprise.
Covered in Projects: shared context for recurring work
3.Anything uploaded in one chat inside a Project is automatically available to the Project's other chats.Why is that wrong?
Only content in the project knowledge base is shared across the Project's chats. Files uploaded inside a single chat stay in that chat.
Covered in Projects: shared context for recurring work
4.Artifacts are a paid-plan feature.Why is that wrong?
Every plan, including Free, can create artifacts in a chat. Only templates (Design, Slides, Docs) are limited to paid plans.
Covered in Artifacts: durable, editable deliverables
Sources
Every claim above is drawn from one of these pages, quoted as it was written on the date shown.
- 1.https://support.claude.com/en/articles/11095361-when-should-i-use-web-search-extended-thinking-and-researchOfficial docs
“Web search is best for straightforward, factual queries that can be answered with one or two tool calls”
↩︎ Start from chat: the quick-answer baseline“Extended thinking on its own is best if you don't want to search the web or use integrations”
↩︎ Start from chat: the quick-answer baseline“The research feature can synthesize information across multiple sources with proper citations and produces a longer final report.”
↩︎ Research: agentic, multi-source investigation with citations“Research is optimal for comprehensive information gathering requiring five or more tool calls over 1-3 minutes”
↩︎ Key concept“It can do things like compare business competitors, update outdated internal documents with info from the web”
↩︎ Checkpoint - 2.
“Claude operates agentically, conducting multiple searches that build on each other while determining exactly what to investigate next.”
↩︎ Research: agentic, multi-source investigation with citations“With research, Claude delivers thorough answers in minutes, complete with easy-to-check citations”
↩︎ Research: agentic, multi-source investigation with citations“such as Gmail, Google Calendar, and Google Docs when connected”
↩︎ Research: agentic, multi-source investigation with citations“research sessions can use up your limits faster due to Claude retrieving multiple sources and providing comprehensive responses”
↩︎ Research: agentic, multi-source investigation with citations“Research is available for users with paid Claude plans (Pro, Max, Team, or Enterprise)”
↩︎ Exam trap 1“Research is available for users with paid Claude plans (Pro, Max, Team, or Enterprise)”
↩︎ Prediction“You must have web search turned on for research to function.”
↩︎ Checkpoint - 3.
“Projects allow you to create self-contained workspaces with their own chat histories and knowledge bases.”
↩︎ Projects: shared context for recurring work“Enhanced project knowledge with RAG is only available to users with paid Claude plans (Pro, Max, Team, or Enterprise).”
↩︎ Projects: shared context for recurring work“For users on Team and Enterprise plans, projects can be shared with other members of your organization”
↩︎ Projects: shared context for recurring work“Projects are available to all users, including those with free Claude accounts. Free users can create a maximum of five projects.”
↩︎ Exam trap 2 - 4.
“Give your project a name and description (note that Claude will not have access to these details).”
↩︎ Projects: shared context for recurring work“Anything you upload to this space will be used across all of your chats within that project.”
↩︎ Exam trap 3“Context is not shared across chats within a project unless the information is added into the project knowledge base.”
↩︎ Checkpoint - 5.https://support.claude.com/en/articles/11473015-retrieval-augmented-generation-rag-for-projectsOfficial docs
“expand your project's capacity by up to 10x while maintaining quality responses”
↩︎ Projects: shared context for recurring work“RAG works with all Claude features, including web search, extended thinking, and Research.”
↩︎ Putting it together: picking the feature for the job - 6.https://support.claude.com/en/articles/9487310-what-are-artifacts-and-how-do-i-use-themOfficial docs
“It's something you're likely to want to edit, iterate on, or reuse outside the conversation.”
↩︎ Artifacts: durable, editable deliverables“Artifacts are available on Free, Pro, Max, Team, and Enterprise plans”
↩︎ Artifacts: durable, editable deliverables“Edit an earlier message to create a different version of the chat, with its own artifacts”
↩︎ Artifacts: durable, editable deliverables“Everything you make is saved to the Artifacts tab in your Claude sidebar”
↩︎ Artifacts: durable, editable deliverables“To start from a template, edit, or change sharing settings, use Claude on the web or Claude Desktop.”
↩︎ Artifacts: durable, editable deliverables“usage counts against each person's own plan limits rather than yours”
↩︎ Artifacts: durable, editable deliverables“Templates (Claude Design, Claude Slides, and Claude Docs) are in beta on paid plans only.”
↩︎ Exam trap 4“Artifacts require Cloud code execution and file creation to be turned on in Settings > Capabilities”
↩︎ Checkpoint - 7.
“which brings Claude Code's agentic capabilities to knowledge work beyond coding”
↩︎ Putting it together: picking the feature for the job“Claude Cowork is available on paid plans (Pro, Max, Team, Enterprise).”
↩︎ Putting it together: picking the feature for the job“Ask for what you need, and Claude decides whether that's a quick answer or a task.”
↩︎ Putting it together: picking the feature for the job