Perplexity Spaces vs ChatGPT Projects vs Claude Projects
Three workspace features solving the same problem from different philosophies. Where each one's design shows through, and which kind of work each is built for.
All three AI tools shipped a "project" or "space" feature in the last 18 months. They look similar. They are not.
All three solve the same problem from different starting assumptions, and the assumption is what shows through in use.
What each one is
Perplexity Spaces. A persistent research workspace. You can add files, focus the AI on specific domains, and keep the context across many sessions. Optimized for research and source-anchored answers.
ChatGPT Projects. A folder for related conversations. You can pin instructions and reference files. Conversations within a project share context. Optimized for ongoing work where you want consistent voice and access to your reference material.
Claude Projects. Same general shape as ChatGPT Projects. You can add custom instructions and reference files. Claude reads the files when relevant.
Where Perplexity Spaces wins
Research workflows. When the workspace is about finding and synthesizing external information, Perplexity's source-anchored architecture is clearly the best.
Domain focus. You can restrict a Space to specific sources (academic, news, your uploaded docs). This filters out the noise more effectively than the alternatives.
Multi-day research projects. The Space holds the cumulative context of all your research. Coming back days later, your prior threads inform the new ones.
Where ChatGPT Projects wins
Writing projects with reference material. If you have a style guide, prior outputs you want to match, brand voice docs — ChatGPT Projects holds them as reference and applies them across conversations.
Multi-conversation organization. You can have many related conversations within a project. Browsing back through is straightforward.
The UX. ChatGPT's interface for managing files and instructions within projects is the most polished of the three.
Where Claude Projects wins
Coding projects. On code as reference material, Claude reads it more accurately than ChatGPT. Conversations about complex codebases feel cleaner.
Long technical references. Claude's longer context window means projects with substantial reference docs work better.
Following detailed custom instructions. Claude tends to follow project-level instructions more reliably across many conversations.
What each one does worse
Perplexity Spaces is weaker on writing. It's a research tool. Writing within a Space feels secondary to the source-finding. Don't try to use it as a writing workspace.
ChatGPT Projects is weaker on long reference material. Past a certain document length the AI's effective use of project files degrades. Trim aggressively.
Claude Projects is the simplest and sometimes the most limited. Fewer features than competitors. Whether that's a bug or a feature depends on you.
Where it fits
Perplexity Spaces suits a multi-week research project. New client industry research. Competitive scans. Anything where a corpus of sources accumulates over time.
ChatGPT Projects suits ongoing writing that has to hold a consistent voice. My newsletter writing has a Project with style guidelines and prior posts.
Claude Projects suits ongoing technical work. My prometheus consulting project has a Project with the codebase reference, brand voice doc, and pillar guidelines.
Different tools for different jobs. The "best Project feature" doesn't exist. The best Project feature for YOUR work depends on what the work is.
Whether to pay for all three
If you do research, writing, and technical work: yes, paying for all three is reasonable. $60/month total. Cheaper than a single hour of an analyst's time.
If you do mostly one kind of work: pick the one that fits your work. Save the money.
The thing nobody mentions
The "Project" feature is each company's bet on the future shape of AI usage. Persistent context, file-based reference, multi-conversation continuity.
The future is here already. The shape is converging. In 12 months these will look more similar than they do today.
The choice today is less about features and more about which company's broader AI tools you want to be embedded in. Perplexity for research-first workflows. ChatGPT for general use. Claude for technical and long-context work.
Where to start
Pick the AI tool you use most often. Use ITS Project feature. The integration with the rest of your workflow matters more than the marginal feature differences between Project implementations.
If you don't have a primary AI tool yet, default to Claude Projects for technical/writing work or Perplexity Spaces for research-heavy work.
Get the next one.
Field notes from real deployments, written the way this one was. No schedule promised and nothing sold to you. Unsubscribe whenever.