Claude Projects + Artifacts Just Went Free: What You Can Actually Do With Them

Enhanced Human·March 14, 2026

Anthropic just made two of Claude's most powerful features free. Until March 11, 2026, Projects and Artifacts were locked behind the $20/month Pro plan. Now anyone with a free Claude account can use them.

If you've been using Claude as a basic chatbot, you've been missing the point. These two features transform it from a question-answering tool into an organized workspace. Here's what changed and whether it's worth your time.

What Projects and Artifacts Actually Do

Projects let you organize conversations around specific work with persistent context. Instead of losing all your conversation history every time you start fresh, Projects remember what you're working on.

Artifacts create a dedicated workspace next to your chat where Claude can generate and edit code, documents, diagrams, and designs. Think of it as a split-screen where the conversation stays on the left and the actual work appears on the right.

Together, they solve two annoying problems: Claude forgetting what you told it three conversations ago, and having to copy-paste outputs into other tools to actually use them.

Who This Is For

This is a work pillar tool. It's built for people managing multiple AI-assisted projects who need:

Separate contexts for different clients or projects
Reusable templates and prompts
Live editing of documents, code, or designs without leaving the chat
A way to pick up where you left off without re-explaining everything

If you only use AI occasionally for one-off questions, Projects are overkill. Stick with the basic chat.

If you're juggling multiple work streams and constantly re-prompting Claude with the same background info, this will save you hours.

Getting Started: Step-by-Step Setup

Creating Your First Project

Go to claude.ai and log in (free account works)
Click the folder icon in the left sidebar or the "+" next to "Projects"
Name your project (be specific: "Q1 Marketing Copy" not "Marketing")
Add instructions in the Project Knowledge section

Project Knowledge is where the magic happens. This is persistent context Claude references in every conversation within that Project. Add:

Your role and goals for this project
Brand voice guidelines
Reference documents (up to 5 files, each up to 200,000 tokens)
Templates you reuse
Common abbreviations or project-specific terminology

Example Project Knowledge for a freelance writer:


You're helping me write B2B SaaS blog posts for [Client Name].

Tone: Conversational but authoritative. Short paragraphs. No jargon.
Audience: Marketing managers at 50-500 person companies.
Goal: 1,200-1,500 word posts optimized for [target keyword].

Never use: "game-changer", "revolutionary", "seamless", em dashes.
Always include: Real examples, data when available, actionable tips.

Now every conversation in that Project automatically has this context. You don't re-explain it.

Using Artifacts

Artifacts appear automatically when Claude generates something interactive or structured. You don't need to enable them - they just show up.

Claude will create an Artifact when you ask it to:

Write or edit code (HTML, CSS, JavaScript, Python, etc.)
Create documents (formatted text, outlines, templates)
Generate diagrams (flowcharts, org charts, system maps using Mermaid syntax)
Build interactive components (calculators, quizzes, simple tools)

Example: "Create an HTML landing page for my coaching service with a hero section, benefits list, and contact form."

Claude writes the code in the left chat. The rendered page appears in the right Artifact panel. You can preview it live, copy the code, or ask Claude to edit it ("Make the hero image full-width" or "Change the button color to green").

You can also edit Artifacts directly. Click into the code or document, make changes, and they persist. Claude sees your edits and can work with them.

Switching Between Projects

Click the folder icon to see all your Projects. Each maintains its own conversation history and context. You can have dozens running simultaneously without crosstalk.

Use cases:

Separate Projects for each client
One for work, one for personal learning
Project per content pillar (if you run a newsletter)
Different coding projects in different languages

Key Features with Real Examples

1. Persistent Context Across Conversations

Before Projects: "I'm writing a blog post about AI productivity tools for small business owners. The tone should be friendly and practical, around 1,200 words, with 3-5 actionable tips..."

You type this every single time you start a new chat.

With Projects: You type it once in Project Knowledge. Every conversation starts with that context already loaded.

Test: I created a "Blog Writing" Project with voice guidelines, added them to Project Knowledge, then started three separate chats asking for post ideas, outlines, and drafts. Claude referenced the guidelines in all three without me repeating them.

2. Live Document Editing in Artifacts

Example: "Write a cold email template for my web design service."

Claude generates the email in an Artifact. You can:

Edit it directly in the panel
Ask Claude to revise specific parts ("Make the subject line shorter")
Copy the final version when ready

The Artifact stays editable. You're not stuck with Claude's first draft.

I tested this by asking Claude to write a landing page, then requested 5 different revisions (color scheme changes, layout tweaks, adding a pricing section). Each change appeared instantly in the live preview. Way faster than copy-pasting into CodePen.

3. Code Generation and Iteration

Example: "Build a JavaScript countdown timer for my product launch. 7 days from now, bold red text, centered on the page."

Claude writes the HTML/CSS/JS in an Artifact. The timer runs live in the preview panel. You see exactly what you're getting.

Then: "Add a 'Notify Me' button below the timer that opens an email signup form."

Claude adds the button and form. The preview updates. You didn't leave the conversation.

For non-developers, this is huge. You can build functional tools (calculators, forms, simple apps) by describing what you want and iterating in plain English.

4. Diagram and Flowchart Generation

Example: "Create a flowchart showing my customer onboarding process: sign up, email verification, profile setup, first project creation, success."

Claude generates a Mermaid diagram in an Artifact. It renders as an actual flowchart, not just code.

Useful for:

Process documentation
System architecture diagrams
Decision trees
Org charts

I tested this for a content workflow diagram (idea > research > draft > edit > publish). Claude built it in 10 seconds. Normally I'd spend 20 minutes in Lucidchart.

5. Multi-File Project Management

You can upload up to 5 files per Project (PDFs, CSVs, text files, code). Claude reads them and references them in every chat.

Example use case: Upload your brand guidelines PDF, a competitor analysis, and last quarter's performance report. Now every conversation in that Project has access to all three without you pasting excerpts.

Limitation: 5 files max, no images, each file capped at ~200,000 tokens (roughly 150,000 words).

Pricing Breakdown

Free Tier (what you get now):

Unlimited Projects
Full Artifacts access
Up to 5 files per Project
Claude 3.5 Sonnet model
Standard rate limits (usage caps during high demand)

Claude Pro ($20/month):

Everything in free
5x higher usage limits
Priority access during peak times
Access to Claude Opus (more powerful model for complex tasks)
Early access to new features

For most users, the free tier is enough. You only hit rate limits if you're having 50+ long conversations per day.

Upgrade to Pro if:

You use Claude all day for work and hit rate limits
You need Opus for highly technical or creative tasks
You want guaranteed access during launches or outages

What It's Bad At

No mobile app editing. You can view Artifacts on mobile (claude.ai works in browsers), but editing is clunky. This is a desktop-first tool.

File upload limits are strict. 5 files per Project, no images, no video. If you're working with visual assets or large datasets, you'll need another tool.

No collaboration. You can't share a Project with teammates. It's single-user only. If you need multi-person access, look at ChatGPT Team or Google's Gemini workspace tools.

Artifacts don't export cleanly. You can copy code or text, but there's no "Export as ZIP" or "Download project files" button. For complex multi-file projects, you'll still need a proper code editor.

Context limits still apply. Projects remember your instructions, but individual conversations still have token limits. Very long chats eventually forget earlier parts of the thread (though Projects help by reducing how much context you need to provide upfront).

How It Compares to ChatGPT Projects

ChatGPT has had Projects since late 2024 (Pro users only). Here's how they stack up:

Feature	Claude Projects (Free)	ChatGPT Projects (Pro $20/mo)
Price	Free	$20/month required
Persistent Context	Yes, up to 5 files	Yes, up to 10 files
Live Editing	Yes (Artifacts)	Limited (Code Interpreter shows output, not live preview)
Diagram Generation	Yes (Mermaid, renders live)	Yes (generates images via DALL-E, not editable diagrams)
Code Execution	No (shows code, you run it)	Yes (Code Interpreter runs Python)
File Types	Text, PDF, CSV, code	Text, PDF, CSV, images, data files
Collaboration	No	No
Web Access	No	Yes (ChatGPT can browse)

Claude wins on: Free access, live code previews, editable diagrams, cleaner document generation.

ChatGPT wins on: Running code directly, handling images, web browsing within Projects, larger file limits.

Verdict: If you're doing writing, design, or front-end coding, Claude's Artifacts are better. If you need data analysis, image generation, or web research, ChatGPT Projects have the edge.

For most non-technical users doing work tasks (emails, blog posts, landing pages, process docs), Claude's free tier beats ChatGPT's $20/month offering.

Verdict: Who Should Use This

Use Claude Projects + Artifacts if you:

Manage multiple clients or work streams and need separate AI contexts
Write a lot (blog posts, emails, marketing copy, reports)
Need to generate and iterate on code, documents, or diagrams quickly
Want organized AI assistance without paying $20/month
Prefer seeing live previews of outputs instead of copy-pasting into other tools

Stick with basic ChatGPT if you:

Only use AI occasionally for one-off questions
Need image generation or web browsing in your Projects
Require data analysis with code execution (Python, data viz)
Work heavily with images or visual assets

Skip both if you:

Don't have recurring projects that need persistent context
Prefer specialized tools for specific tasks (Grammarly for writing, Cursor for coding, Whimsical for diagrams)

Work Pillar Notes: Integration and Workflow

This is a work tool built for context organization, not task automation. It won't schedule your meetings or send emails for you.

Where it excels:

Content creation workflows: Separate Projects for each client, brand voice in Project Knowledge, Artifacts for drafts
Freelance developers: Project per client, starter code templates in Knowledge, live previews via Artifacts
Consultants: Project per engagement, reference docs uploaded, conversation history preserved
Marketing teams: Campaign-specific Projects, brand guidelines in Knowledge, ad copy / landing page generation

Best practice: Don't create a Project until you have recurring work in that area. One-off tasks belong in regular chats. Projects are for things you return to weekly.

Integration gaps: No Slack, no Google Docs, no API access on the free tier. This is a standalone workspace. You copy outputs into your production tools.

Getting the Most Out of It

Write detailed Project Knowledge. The more specific your instructions, the better Claude performs. Include examples, not just rules.

Use descriptive Project names. "Client: Acme Corp - Blog Posts" beats "Acme" when you're managing 10 Projects.

Leverage Artifacts for iteration. Don't ask Claude to "write a final draft." Ask for a draft, review it in the Artifact, then request specific changes. The live editing makes iteration fast.

Upload reference docs. If you refer to the same guidelines, data, or examples repeatedly, upload them once instead of pasting excerpts every chat.

Create template Projects. Set up a "Blog Post Template" Project with your standard instructions, then duplicate it for each new client (when Anthropic adds that feature - for now, recreate manually).

Bottom Line

Claude Projects and Artifacts going free is a big deal. You're getting organizational tools that cost $20/month elsewhere, with better live editing than most alternatives.

If you use AI for work more than a couple times a week, set up at least one Project today. You'll immediately notice the difference between "starting from scratch every chat" and "picking up where you left off."

The free tier has limits, but they're generous. You'd need to be a very heavy user to need Pro.

Try it for a week with one work Project. If you're not using it by day 3, you probably don't need it. If you find yourself creating more Projects by day 7, you've found a new core tool.

Get started: https://claude.ai (free account, no credit card required)

All articles