This is a research-based review. It's built from Perplexity's official pricing and product pages plus public reporting — not from a hands-on test in our own account. Where something is widely reported rather than confirmed by us, we say so. If we test Perplexity directly later, we'll update this and relabel it.
Perplexity is built around a single idea: a question deserves a cited answer, not ten search results. You ask, it searches the web, synthesises across sources, and returns a written response with footnotes you can click through to verify. In 2026 it's grown into a four-product platform — the search engine, the Comet browser (now free), Spaces for collaborative research collections, and the Sonar API for developers — with a pricing ladder that mixes a generous free tier, a $20 Pro plan, and a $200 Max plan aimed at heavy research users.
Perplexity at a glance
The plans
Pro at $20 is the right call for most serious users. You get unlimited Pro Searches, model switching across Claude Sonnet 4.6, GPT-4o, Mistral Large and Perplexity's own Sonar models, and Comet Plus (premium publisher content in the browser) bundled in. The bigger lift is Max, which adds Claude Opus 4.6, GPT-5.2, and a feature called Model Council: launched in February 2026, it sends your query to three frontier models at once (Opus 4.6, GPT-5.2, Gemini 3 Pro) and synthesises a response that surfaces where the models agree, disagree, and add unique insights.
Comet went free in 2026
The biggest 2026 change for casual users is that Comet — Perplexity's AI browser — dropped the paywall on March 18, 2026 and rolled out free on iOS, Android, Windows and Mac. Free Comet ships with the core AI features intact: agentic search, page summarisation, voice mode, shopping assistance, and Deep Research from inside the browser. The paid extra is Comet Plus at $5/mo, which unlocks premium publisher content (and is bundled free with Pro and Max). If you've never tried Perplexity, the free browser is now the easiest on-ramp.
Our score breakdown
Scores reflect documented capability and public reception, not a hands-on test in our account. Breadth is marked down by design — Perplexity is search-first, not a general assistant.
Pros & cons
Who should use Perplexity?
Use it if a big share of your AI use is "what's the current state of…", "what do these sources say about…", "compare X and Y", or "find me the latest research on…". Cited, current answers are the whole product, and Pro at $20 makes the model switching and unlimited searches worth it for anyone who runs a few research questions a day.
Look elsewhere if you want a general-purpose chat assistant for writing, thinking, code, and tool use — that's ChatGPT or Claude territory. Many serious users keep Perplexity for research and one of those for the rest.
8.6/10. The best AI search engine in 2026, and at $20/mo Pro is the right tier for almost anyone who pays. The 2026 changes — free Comet, model switching, Model Council — make it a stronger value than it was a year ago. Just treat it as the research tool in your stack, not the only tool.