Most people who use Perplexity AI every day have never touched the Perplexity AI Focus Modes feature. They type their question, leave the search source on its default setting, and read whatever comes back. What they are missing is the single most powerful tool for improving answer quality on the platform. Perplexity AI Focus Modes let you control exactly where the AI searches for information before it generates a response. The difference between using the right mode and ignoring it is the difference between getting an answer grounded in peer-reviewed research and getting one grounded in blog posts. This guide covers every mode available in 2026, when to use each one, real prompt examples, and the specific use cases where each mode outperforms the default.

What Are Perplexity AI Focus Modes?

Perplexity AI Focus Modes are source filters that restrict where the platform searches for information before it writes a response to your query. When you select a Focus Mode, you are not just changing a cosmetic setting. You are telling the AI’s retrieval system to ignore certain types of sources entirely and pull only from the category you have chosen. The result is an answer that is more precisely matched to the type of information you actually need.

By default, Perplexity AI searches the entire open web for every query. This works well for general questions, current events, and everyday research. It is a poor choice, however, for academic research where you need peer-reviewed sources, for mathematical problem-solving where you need computational precision, or for understanding public opinion on a topic where social discussions are far more relevant than formal articles.

Perplexity AI Focus Modes solve each of these problems by routing your query through a specialised source layer before any text is generated. The mode you select determines which knowledge network Perplexity searches. Academic Mode locks it to scholarly databases. Social Mode locks it to forums and community discussions. Math Mode connects it to Wolfram Alpha for computational results. Writing Mode removes live search entirely and uses the model’s internal knowledge for composition tasks. Finance Mode restricts results to financial data sources and SEC filings.

Understanding which Perplexity AI Focus Mode fits each task makes you a dramatically more effective user. Most casual users leave the setting on Web permanently. Power users switch modes deliberately for every research session and consistently get more accurate, more relevant, and more credible results.

How Perplexity AI Focus Modes Work?

Pre-Search Source Filtering

The key thing to understand about Perplexity AI Focus Modes is that the filtering happens before the AI generates a response, not after. When you select Academic Mode, for example, Perplexity does not search the whole web and then filter the results for academic content. It initialises a different retrieval pathway that exclusively queries scholarly databases, journals, and academic repositories. The reasoning tokens the model uses are dedicated to the logic of that specific source type from the start of the session.

This architectural detail matters because it explains why Perplexity recommends setting your Focus Mode before starting a thread rather than changing it midway through. Switching modes mid-conversation requires the system to re-index previous messages under a different retrieval logic, which is why it currently requires starting a fresh thread for consistency and accuracy. For best results, decide on your mode before you type your first query.

How Source Restriction Improves Answer Quality

When Perplexity AI searches the entire web without a focus filter, it pulls from a mix of authoritative sources, SEO-optimised blog content, news articles of varying quality, forum posts, and commercial pages. For most general queries, this mix is fine. For specialised research, it introduces noise that reduces the reliability of the answer.

Selecting the right Perplexity AI Focus Mode eliminates that noise at the source level. An Academic Mode query on a medical topic pulls from peer-reviewed journals rather than health blogs. A Social Mode query on a product pulls from genuine user discussions rather than brand-managed review pages. A Math Mode query on a calculus problem runs through Wolfram Alpha’s computational engine rather than a tutorial website that may have errors. The answer quality improvement is not subtle. It is consistent and measurable across repeated use.

Every Perplexity AI Focus Mode Explained

Web Mode (Default)

Web Mode is the default setting for all Perplexity AI searches. It queries the entire open internet, pulling from news sites, blogs, reference pages, forums, academic sites, and commercial pages based on relevance signals. The answer synthesises the most relevant content from across all of these source types simultaneously.

Best for: General research, current events, product comparisons, everyday factual questions, trending topics, news, and any query where a broad range of sources is appropriate.

When not to use it: When you specifically need peer-reviewed sources, computational precision, or community opinions separated from formal content. Web Mode treats all source types equally, which is its strength for general queries and its weakness for specialised ones.

Real prompt example: “What are the best lightweight laptops for students available in India under ₹60,000 in 2026?”

Academic Mode

Perplexity AI Focus Modes does not have a more important individual setting than Academic Mode for research-intensive users. When selected, it restricts retrieval entirely to scholarly databases, peer-reviewed journals, academic publications, pre-print repositories like arXiv, and similar high-authority academic sources. The citations it returns include author names, publication titles, journal names, and publication dates linked directly to the original papers.

Academic Mode effectively tells Perplexity to ignore any source that has not undergone formal review or publication processes. For university-level research, evidence-based professional decisions, and any context where the credibility of the source matters as much as the content, this mode is the correct choice.

Best for: Literature reviews, university essays, understanding the state of research on any topic, fact-checking scientific claims, medical research, legal research with academic grounding, and any query where you need to cite or verify scholarly sources.

When not to use it: For current events, practical how-to questions, or any topic where the most relevant information is not in academic literature. Academic Mode will produce poor results for questions like “What is the best free project management tool?” because that is not the kind of question academic databases address.

Real prompt example: “What does current peer-reviewed literature say about the effectiveness of spaced repetition for long-term vocabulary retention in second language acquisition? Focus on studies published after 2021.”

Social Mode

Social Mode restricts Perplexity AI’s search to social media platforms, online forums, discussion boards, and community-driven content including Reddit, Quora, and similar platforms. The answers it produces reflect real-time public opinion, community experience, and user-generated perspectives rather than formal editorial content.

This mode is particularly valuable when you want to understand what real users think about a product, service, or experience. Review sites can be manipulated. Brand websites present curated information. Community discussions on Reddit or product forums tend to reflect genuine, unfiltered user sentiment, both positive and negative.

Best for: Understanding real user opinions on products and services, finding community-recommended solutions to practical problems, gauging public sentiment on topics, researching user experiences that would not appear in formal reviews, and finding practical advice from people who have actually done what you are asking about.

When not to use it: For factual research, academic topics, financial data, or any context where verified accuracy matters more than community opinion. Social Mode surfaces discussions that can contain misinformation, outdated advice, or highly subjective accounts.

Real prompt example: “What do actual users on Reddit and forums say about using Notion versus Obsidian for personal knowledge management? What are the most common complaints about each?”

Writing Mode

Writing Mode removes live web search entirely. When selected, Perplexity AI uses the internal knowledge of its language model to assist with composition, editing, structuring, grammar, and creative writing tasks without pulling from any external sources. This is appropriate when you are not looking for researched information but for help with the craft of writing itself.

Because there is no live retrieval, Writing Mode is faster than other modes and does not produce citations. The trade-off is that the content is drawn from training data rather than current web sources, so it is not appropriate for any task requiring up-to-date or externally verified information.

Best for: Drafting emails, essays, blog posts, social media copy, cover letters, creative writing, editing and improving existing text, restructuring documents, rephrasing sentences, and any task where writing quality is the goal rather than information retrieval.

When not to use it: For research questions, current events, factual queries, or anything where you need sourced and verifiable information. Using Writing Mode for research will produce confident-sounding answers with no citations and no way to verify the claims.

Real prompt example: “Help me write a professional email declining a job offer while expressing genuine appreciation and keeping the door open for future opportunities.”

Math Mode

Math Mode connects Perplexity AI to Wolfram Alpha’s computational engine, which is one of the most accurate mathematical and scientific calculation systems available. Instead of generating a written explanation of how to solve a problem, Math Mode computes the answer directly using structured mathematical logic, showing intermediate steps in a structured proof format.

This is significantly more accurate for mathematical problems than standard AI responses, which can generate plausible-looking but incorrect calculations. The integration with Wolfram Alpha means computational tasks are handled by a purpose-built calculation system rather than a language model estimating a result.

Best for: Solving equations, calculus problems, algebra, statistics, physics calculations, unit conversions, mathematical proofs, data analysis requiring precise computation, and any quantitative problem where computational accuracy is essential.

When not to use it: For conceptual explanations of mathematical topics at a general level, writing mode or web mode may produce more readable explanations. Math Mode is optimised for computation, not for explaining mathematical concepts in plain language for beginners.

Real prompt example: “Solve the differential equation dy/dx = 3x² + 2x, given that y = 5 when x = 0. Show all steps.”

Video Mode

Video Mode restricts Perplexity AI’s search to YouTube and returns answers that draw from video content rather than text-based web pages. This is useful when a visual or tutorial format is the most relevant type of source for your question, or when you want to understand what video content exists on a topic before deciding what to watch.

Perplexity AI does not play videos within the interface. Video Mode finds relevant YouTube content, summarises it, and cites the specific videos so you can click through to watch them. This is particularly useful for technical tutorials, product demonstrations, how-to content, and educational topics where video explanations are often clearer than written ones.

Best for: Finding video tutorials, understanding what YouTube content exists on a topic, getting summaries of relevant videos before committing time to watch them, product demonstrations, technical walkthroughs, and learning topics that benefit from visual explanation.

Real prompt example: “What are the best YouTube tutorials for learning Figma from scratch in 2026? Summarise what each covers.”

Finance Mode

Finance Mode restricts Perplexity AI’s search to financial data sources, SEC filings, annual reports, corporate disclosures, and financial news publications. It is designed for retrieving and analysing financial information with a level of source specificity that Web Mode cannot match for this type of query.

When you ask a financial question in Finance Mode, Perplexity prioritises primary financial documents and authoritative financial data over general commentary. This makes it the appropriate mode for investors, financial analysts, and anyone conducting due diligence on a company or market.

Best for: Researching company financials, reading SEC filings, understanding corporate disclosures, tracking stock performance data, analysing market trends with financial data support, and any research task that depends on primary financial documents rather than general news coverage.

Real prompt example: “What were the key financial highlights from Tesla’s most recent SEC annual filing, including revenue, net income, and major risk factors disclosed?”

How to Switch Between Perplexity AI Focus Modes?

Switching between Perplexity AI Focus Modes takes three seconds and one click. Here is the process:

Open perplexity.ai and start a new thread. Before typing your query, locate the Focus button below or beside the main search bar depending on whether you are on desktop or mobile. Click it. A dropdown menu appears showing all available modes: Web, Academic, Social, Writing, Math, Video, and Finance. Click the mode that fits your query. The search bar updates to reflect your selection. Type your query and submit.

The Focus selection applies to all queries within that thread. If you need to switch modes for a different type of question, start a new thread and select the appropriate mode before submitting. Changing the mode mid-thread is not currently supported and requires a fresh start.

On mobile, the Focus button appears in the same position relative to the search bar. The mode selection works identically on the iOS and Android apps. All Focus Modes are available on all plans including the free tier, which means you do not need a Pro subscription to use Academic Mode, Social Mode, or any other mode.

Key Benefits of Using the Right Focus Mode

Dramatically Higher Answer Relevance

The most consistent benefit of using Perplexity AI Focus Modes correctly is that the answers match your actual research need more precisely. A student using Academic Mode for a literature review receives answers grounded in peer-reviewed publications rather than SEO-optimised health blogs or Wikipedia summaries. A marketer using Social Mode to research consumer sentiment receives community-sourced perspectives rather than brand press releases. The mode does not change how Perplexity writes an answer. It changes what Perplexity reads before writing it. That upstream change has a substantial downstream effect on the quality and credibility of what you receive.

Reduced Noise From Irrelevant Sources

The open web contains enormous amounts of content that is optimised for visibility rather than accuracy. SEO-driven blog posts, affiliate marketing content, and promotional pages can appear at the top of web searches on topics like product reviews, health information, and financial advice. Perplexity AI Focus Modes filter these out at the source level. When Academic Mode is active, that entire category of content is invisible to the retrieval system. The model never sees it and cannot cite it. For users who have experienced frustration with AI tools confidently summarising low-quality sources, Focus Modes remove the root cause rather than treating the symptom.

Faster, More Targeted Research Workflows

Using the right Perplexity AI Focus Mode reduces the number of follow-up questions needed to get a useful answer. A Web Mode answer to an academic question often requires follow-up prompts to narrow the sources, add scholarly context, or remove anecdotal content. An Academic Mode answer to the same question arrives already grounded in the right source type. Over the course of a research session involving multiple queries, this reduction in iteration time compounds into a meaningfully faster workflow. For professionals and students who conduct research regularly, this is one of the most practical efficiency improvements available in the platform without any additional cost.

No Additional Cost, Available on All Plans

All Perplexity AI Focus Modes are available on every plan including the free tier. There is no paywall separating Academic Mode from Web Mode. Any user who has created a free account can access every focus filter immediately. This means the performance improvement that Focus Modes provide is available to every user without any financial commitment. Given that most users leave the default Web Mode active permanently, simply learning which mode fits each task represents a meaningful quality upgrade with zero cost.

Perplexity AI Focus Modes: Quick Reference Table

Focus Mode	Sources Searched	Best Use Case	Available On
Web	Entire open internet	General research, current events, everyday questions	All plans
Academic	Peer-reviewed journals, scholarly databases, arXiv	University research, literature reviews, evidence-based queries	All plans
Social	Reddit, forums, community discussions	User opinions, product experiences, public sentiment	All plans
Writing	Internal model knowledge (no live search)	Drafting, editing, creative writing, composition	All plans
Math	Wolfram Alpha computational engine	Equations, calculus, statistics, precise computation	All plans
Video	YouTube content	Tutorials, demonstrations, video-based learning	All plans
Finance	SEC filings, financial data, corporate reports	Company research, investment analysis, financial data	All plans

Who Benefits Most from Each Focus Mode?

Academic Mode is built for students, researchers, and evidence-based professionals. If you regularly produce work that needs to be grounded in published research, peer-reviewed literature, or verifiable scholarly sources, Academic Mode changes the character of every query you run. Undergraduate and postgraduate students writing essays, scientists reviewing literature, doctors researching clinical evidence, and policy analysts looking for research-backed data all belong in Academic Mode for their core research work. For a detailed guide on using this mode as part of a full student research workflow, see our Perplexity AI for students guide.

Social Mode is built for marketers, product researchers, and community managers. Anyone whose work depends on understanding what real people think, experience, and discuss about a product, service, or topic will get more honest and unfiltered data from Social Mode than from Web Mode. Consumer sentiment research, competitor experience analysis, and community trend identification are all better served by Social Mode’s focus on forum and discussion content.

Math Mode is built for students, engineers, scientists, and data analysts. Any quantitative problem that requires accurate computation rather than a written explanation belongs in Math Mode. Students working through problem sets, engineers verifying calculations, and data analysts running quick statistical checks should use Math Mode for numerical queries rather than trusting Web Mode to produce accurate computations from a language model.

Writing Mode is built for content creators, professionals, and anyone drafting structured text. When the task is composition rather than research, Writing Mode removes the distraction of live web retrieval and lets the model focus on language, structure, and clarity. Bloggers, marketers, executives drafting communications, and students writing creative or reflective pieces will get more focused writing assistance from Writing Mode than from a mode that is simultaneously trying to search the web.

Finance Mode is built for investors, analysts, and business researchers. Anyone conducting due diligence, tracking corporate financial performance, or researching market data should use Finance Mode to ensure the answers are grounded in primary financial documents rather than general commentary. The specificity of the source restriction is what makes it valuable for this audience.

Frequently Asked Questions

Are Perplexity AI Focus Modes available on the free plan?

Yes, all Perplexity AI Focus Modes are available on every plan including the free plan. You do not need a Pro subscription to access Academic Mode, Social Mode, Math Mode, or any other focus filter. The modes are accessible to any user with a free account by clicking the Focus button before submitting a query. The free plan does limit the number of Pro Search queries per day to approximately five, but Focus Mode selection itself is unrestricted. This makes Focus Modes one of the highest-value features available without any payment, and one of the most overlooked by users who have not discovered it.

Can I change Focus Modes in the middle of a conversation?

Not currently. Perplexity AI Focus Modes determine the retrieval pathway and reasoning logic from the beginning of a thread. Changing the mode mid-conversation would require the system to re-index all previous messages under a different source logic, which is not supported in the current interface. If you need to switch from Web Mode to Academic Mode for a follow-up question on the same topic, you need to start a new thread and select the appropriate mode before submitting your first query. This is a deliberate design choice to ensure consistency and accuracy within a session. The practical workaround is to be intentional about mode selection before starting any research thread.

Which Focus Mode is best for university research and assignments?

Academic Mode is the correct choice for the majority of university research. It restricts Perplexity’s search to peer-reviewed journals, scholarly databases, and academic publications, and returns citations that include author names, journal titles, and publication dates. For literature reviews, understanding the state of research on a topic, and finding credible sources for essays, Academic Mode consistently outperforms the default Web Mode. Math Mode is the appropriate addition for any quantitative or computational aspect of your research. Social Mode can be useful for understanding public discourse on a topic as a supplementary research angle, not as a primary academic source. For a full step-by-step guide on using Perplexity for academic work, see our Perplexity AI for students guide.

Does Video Mode let me watch YouTube videos inside Perplexity?

No. Video Mode does not play videos within the Perplexity interface. It searches YouTube content, reads the available text data associated with relevant videos including titles, descriptions, and transcripts where available, and synthesises that content into a written answer. It cites the specific videos it drew from, with links you can click to open them on YouTube. The value of Video Mode is in quickly identifying which YouTube content is most relevant to your query and getting a summary of what each video covers before deciding what to watch. It saves you from manually searching YouTube and clicking through multiple videos to find the most useful one.

How is Math Mode different from asking a math question in Web Mode?

Math Mode connects Perplexity to Wolfram Alpha’s dedicated computational engine, which calculates answers using structured mathematical logic and verifies results computationally. Web Mode uses the language model to generate a mathematical answer from patterns in training data, which can produce plausible-looking but incorrect results, particularly for complex problems. The practical difference is that Math Mode is significantly more reliable for equations, calculus, statistics, and any computation where precision matters. Web Mode may produce an answer that looks correct but contains errors the model cannot detect. For serious mathematical work, Math Mode is not optional. It is the only appropriate mode for quantitative queries on the platform.

Does using Focus Modes count against my daily Pro Search limit?

Focus Mode selection itself does not consume Pro Search queries. Pro Search is a separate feature that activates more thorough multi-step reasoning and retrieval regardless of which Focus Mode is selected. You can run standard searches in any Focus Mode without touching your daily Pro Search quota. If you activate Pro Search in addition to selecting a Focus Mode, that combination counts as one Pro Search query. Free users get approximately five Pro Searches per day. Pro and Max users have effectively unlimited Pro Searches. You can use Focus Modes with standard search as many times as you like without any quota restrictions on any plan.

Final Thoughts

Perplexity AI Focus Modes are the most underused feature on a platform that most users already underuse relative to its full capability. The modes are free, available on every plan, and take three seconds to activate. The improvement in answer quality when you match the right mode to the right task is consistent and immediate. Academic Mode produces genuinely scholarly answers. Social Mode produces genuinely community-grounded opinions. Math Mode produces computationally accurate results. Writing Mode produces focused composition assistance without retrieval noise.

The habit to build is simple: before submitting any query, spend three seconds asking which type of source is actually most relevant to what you need. Then select the corresponding mode. That one habit applied consistently across every research session will produce noticeably better results than leaving the default setting untouched. For the full picture of how Focus Modes fit into the broader Perplexity AI workflow, our complete guide to using Perplexity AI covers every major feature from search to Deep Research. If you are considering upgrading your plan to access Pro Search alongside Focus Modes, our Perplexity AI plans breakdown explains exactly what each tier includes. And if you are deciding between Perplexity and ChatGPT for your research workflow, our Perplexity AI vs ChatGPT comparison covers the key differences in detail. Start with Academic Mode on your next research query and compare the result to what you usually receive. That single test is the fastest way to understand why Focus Modes matter.