Perplexity Keyboard Shortcuts
Perplexity's core distinguishing feature isn't a keyboard shortcut at all — it's that nearly every substantive answer arrives with numbered inline citations linking back to the specific web sources it drew from, a structural difference from a general chat product like Claude.ai or Gemini that treats sourcing as optional context rather than a built-in, always-present part of the response format. That citation-first design is reinforced by Focus modes, a selector that narrows which class of sources Perplexity draws from for a given query — Academic mode leans toward scholarly papers, Writing mode reduces web-search grounding in favor of the model's own generation for creative tasks, and there are several others — which functions more like a search-scope filter than anything a traditional chat app needs, since a plain conversational assistant isn't built around the same underlying source-retrieval step for every single answer. Perplexity also blurs the line between 'search engine' and 'chat assistant' more deliberately than its competitors: the same message box handles a quick factual lookup as comfortably as a longer back-and-forth conversation, and Pro users get access to a more thorough multi-step 'Pro Search' mode that runs several underlying queries and cross-references sources before composing a single answer, taking noticeably longer than a standard response in exchange for more thoroughly checked output. As with the other AI chat interfaces on this site, the keyboard footprint stays close to the near-universal baseline (send, newline, stop generating) since the meaningful differentiation here happens in retrieval and sourcing behavior rather than in dedicated keystrokes, and Focus mode selection, source-list expansion, and switching to Pro Search are all mouse-and-click interactions rather than shortcut-bound ones in the current interface.
Conversation Shortcuts
| Action | Windows | Mac | Description |
|---|---|---|---|
| Send the current message | Enter | Enter | Kicks off the search-and-answer process for whatever you've typed, triggering Perplexity's source retrieval step before the cited response starts appearing. |
| Insert a line break without sending | Shift+Enter | Shift+Enter | Drops in a line break within the same query rather than firing it off right away, the combination to reach for when a multi-part research question needs to stay together as one submission instead of being cut off partway through. |
| Stop response generation | Esc | Esc | Interrupts an in-progress answer, useful if the response is clearly not addressing your actual question or you've already spotted the specific fact you needed among the sources being cited. |
| Start a new thread | Ctrl+I (varies by version) | Cmd+I (varies by version) | Starts a fresh query thread with no prior conversation context carried over, distinct from a follow-up question within an existing thread, which retains the earlier exchange and its cited sources as context. |
Frequently Asked Questions
What's the real functional difference between Perplexity and a general chat assistant like ChatGPT or Claude.ai?
The clearest difference is that Perplexity is built around retrieving and citing live web sources as a core part of nearly every response by default, presenting itself as a research and search tool first, whereas general chat assistants can search the web as an optional capability but aren't structured around inline numbered citations being present in essentially every substantive answer. Perplexity's Focus modes, which scope what kind of sources a given query draws from, don't have a direct equivalent in a general-purpose chat product either.
Is there a keyboard shortcut to switch Focus modes?
No — selecting a Focus mode (such as Academic or Writing) is a mouse-driven interaction through a dropdown or selector near the query box in the current interface, rather than something bound to a keyboard shortcut.
What does Pro Search actually do differently from a standard query?
Pro Search, available to paid Perplexity Pro subscribers, runs multiple underlying searches and cross-references several sources before composing a single answer, rather than retrieving sources once and generating a response immediately. This takes noticeably longer than a standard query but is intended to produce a more thoroughly checked answer for research-heavy questions where a single quick lookup might miss important context or conflicting information.
Can I trust the citations Perplexity provides without checking the original sources myself?
The citations are a genuine improvement over a response with no sourcing at all, since they let you verify a specific claim against its original source directly, but they aren't a substitute for actually reading the linked source when accuracy genuinely matters — a citation confirms where a claim came from, not that the claim itself was interpreted correctly or that the source itself is reliable, so treating citations as a starting point for verification rather than proof in themselves is the safer habit.
Does Perplexity remember earlier threads the way a saved conversation history does in other chat apps?
Yes, past threads are saved to your account and accessible from a history panel, similar to how other AI chat products retain conversation history, though each new thread by default starts without carrying over context from a separate, previous thread unless you explicitly reference it.
Is Perplexity free to use, and what does the paid Pro tier add beyond citations?
Perplexity offers a free tier covering standard cited search-and-answer queries, with the paid Pro subscription primarily unlocking Pro Search's multi-step research mode, a higher usage allowance, and access to choosing among different underlying AI models for a given query, rather than gating the core citation-based answer format itself behind a paywall.
Can Perplexity answer questions about very recent events, given that AI models are typically trained on data with a cutoff date?
Yes, and this is a direct benefit of its retrieval-first design — because Perplexity searches and cites live web sources for most queries rather than relying purely on a model's static training data, it can generally surface and cite genuinely recent information that postdates any individual model's training cutoff, which is a meaningful practical difference from a chat assistant answering purely from memorized training data with no live retrieval step.
Does Perplexity let me pick which underlying AI model answers my query?
Pro subscribers generally have the option to select among several different underlying models for a given query rather than being locked to a single fixed model, which is a meaningful point of difference from a product like Claude.ai or Gemini, where you're interacting with that company's own model family exclusively rather than choosing between models built by different organizations within the same interface.
Is Perplexity's citation-first approach actually more accurate than a general chat assistant, or just more transparent?
It's more accurate to describe it as more verifiable rather than inherently more accurate — the citations let you check a claim against its stated source directly, which is a genuine trust advantage, but a cited source can itself be wrong, outdated, or misread by the summarizing model, so the citations reduce the effort needed to verify a claim without eliminating the need for that verification altogether.