{"id":19346,"date":"2026-09-14T09:00:53","date_gmt":"2026-09-14T13:00:53","guid":{"rendered":"https:\/\/wp.glbgpt.com\/?p=19346"},"modified":"2026-09-14T09:00:54","modified_gmt":"2026-09-14T13:00:54","slug":"perplexity-ai-models-and-modes","status":"publish","type":"post","link":"https:\/\/wp.glbgpt.com\/th\/hub\/perplexity-ai-models-and-modes","title":{"rendered":"\u0e42\u0e21\u0e40\u0e14\u0e25\u0e41\u0e25\u0e30\u0e42\u0e2b\u0e21\u0e14 AI \u0e02\u0e2d\u0e07 Perplexity: \u0e04\u0e38\u0e13\u0e04\u0e27\u0e23\u0e43\u0e0a\u0e49\u0e41\u0e1a\u0e1a\u0e44\u0e2b\u0e19?"},"content":{"rendered":"<style>\n.pplx-text,.pplx-list,.pplx-table,.pplx-heading,.pplx-note,.pplx-code,.pplx-figure,.pplx-toc,.pplx-cta{box-sizing:border-box;font-family:system-ui,-apple-system,\"Segoe UI\",sans-serif;color:#243e40;max-width:100%;}\np.pplx-text{font-size:17px;line-height:1.85;margin:19px 0;}\n.pplx-text a,.pplx-list a,.pplx-table a,.pplx-note a{color:#14756d;text-decoration:underline;text-underline-offset:3px;}\nh2.pplx-heading{font-size:30px;line-height:1.28;letter-spacing:-.6px;margin:54px 0 22px;padding-bottom:15px;border-bottom:1px solid #d9e2dc;scroll-margin-top:28px;}\nh3.pplx-heading{font-size:22px;line-height:1.4;margin:30px 0 12px;scroll-margin-top:28px;}\n.pplx-list{font-size:17px;line-height:1.8;padding-left:26px;margin:20px 0;}.pplx-list li{margin:9px 0;}\nfigure.pplx-table{margin:26px 0;overflow-x:auto;border:1px solid #ceddd7;border-radius:12px;}\n.pplx-table table{border-collapse:collapse;width:100%;min-width:620px;font-size:15px;line-height:1.65;table-layout:auto;}\n.pplx-table th,.pplx-table td{text-align:left;vertical-align:top;padding:16px 18px;border-bottom:1px solid #dce5df;}\n.pplx-table th{background:#dae9e2;color:#193f3c;font-weight:700;}.pplx-table tbody tr:nth-child(even){background:#f2f6f1;}.pplx-table tr:last-child td{border-bottom:0;}\np.pplx-note{background:#faf0db;border-left:4px solid #c99743;border-radius:0 10px 10px 0;padding:21px 24px;font-size:16px;line-height:1.75;margin:26px 0;}\n.pplx-toc{background:#edf4ef;border-radius:12px;padding:24px 28px;margin:30px 0;}.pplx-toc strong{font-size:15px;letter-spacing:1px;text-transform:uppercase;}.pplx-toc ol{display:grid;grid-template-columns:1fr 1fr;gap:13px 28px;padding-left:21px;margin:20px 0 0;}.pplx-toc a{color:#24645c;text-decoration:none;font-size:15px;line-height:1.5;}\n.pplx-code{display:block;background:#eef1ed;border:1px solid #dce2d9;border-radius:10px;padding:22px;white-space:pre-wrap;overflow-wrap:anywhere;font:15px\/1.75 ui-monospace,Consolas,monospace;}\n.pplx-figure{margin:30px 0;}.pplx-figure img{display:block;width:100%;height:auto;border-radius:12px;}.pplx-figure figcaption{color:#657774;font-size:14px;line-height:1.6;margin:12px 4px;}\n.pplx-cta{margin:32px 0;background:#163f42;color:#f9fff6;padding:28px 30px;border-radius:12px;}.pplx-cta p{font-size:17px;line-height:1.75;margin:0 0 18px;}.pplx-cta a{display:inline-block;color:#183e3d;background:#e1eecf;border-radius:6px;padding:12px 19px;text-decoration:none;font-size:15px;font-weight:700;}\n@media(max-width:640px){p.pplx-text,.pplx-list{font-size:16px;line-height:1.8;}h2.pplx-heading{font-size:25px;margin-top:40px;}h3.pplx-heading{font-size:20px;}.pplx-toc{padding:21px;}.pplx-toc ol{grid-template-columns:1fr;gap:11px;}p.pplx-note,.pplx-cta{padding:20px;}.pplx-table th,.pplx-table td{padding:12px;}.pplx-code{font-size:14px;padding:17px;}}\n<\/style>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\"><strong>For most people, the best starting point is Search with Best selected.<\/strong> Perplexity AI models are the engines behind its answers; modes decide whether you get a quick answer, a research report, a lesson, or completed work. Use Thinking for a difficult multi-step question and Deep Research when you need a substantial report with sources.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">The confusing part is that these choices sit alongside subscription tiers. Pro unlocks manual selection of eligible Search models, while Max adds some higher-tier models and features. Changing the model does not automatically change the workflow, and a familiar model name does not give you every feature of its provider\u2019s own app.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">If your work also includes writing, images, and video, GlobalGPT offers an all-in-one workspace with one account and subscription for multiple supported model families. That is useful for consolidating separate AI subscriptions; Perplexity remains the more direct choice when you specifically want its native search and research experience.<\/p>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<figure class=\"wp-block-image size-large\"><img fetchpriority=\"high\" decoding=\"async\" width=\"1024\" height=\"645\" src=\"https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-1024x645.png\" alt=\"\u0e2b\u0e19\u0e49\u0e32\u0e2b\u0e25\u0e31\u0e01\u0e02\u0e2d\u0e07 globalgpt\" class=\"wp-image-19280\" srcset=\"https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-1024x645.png 1024w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-300x189.png 300w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-18x11.png 18w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-767x483.png 767w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-1536x967.png 1536w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-2048x1289.png 2048w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<div class=\"wp-block-buttons is-content-justification-center is-layout-flex wp-container-core-buttons-is-layout-3e41869c wp-block-buttons-is-layout-flex\">\n<div class=\"wp-block-button\"><a class=\"wp-block-button__link has-black-color has-text-color has-background has-link-color wp-element-button\" href=\"https:\/\/www.glbgpt.com\/home?inviter=hub_popup&amp;login=1\" style=\"background:linear-gradient(135deg,rgba(250,183,0,0.55) 0%,rgba(255,106,0,0.66) 96%)\"><strong>\u0e25\u0e2d\u0e07\u0e43\u0e0a\u0e49\u0e41\u0e1a\u0e1a\u0e08\u0e33\u0e25\u0e2d\u0e07\u0e0a\u0e31\u0e49\u0e19\u0e19\u0e33\u0e01\u0e27\u0e48\u0e32 100 \u0e41\u0e1a\u0e1a\u0e1a\u0e19 GlobalGPT<\/strong><\/a><\/div>\n<\/div>\n<\/div>\n\n\n\n<nav class=\"pplx-toc\" aria-label=\"\u0e40\u0e19\u0e37\u0e49\u0e2d\u0e2b\u0e32\u0e02\u0e2d\u0e07\u0e1a\u0e17\u0e04\u0e27\u0e32\u0e21\"><strong>Find your next step<\/strong><ol><li><a href=\"#which-mode\">Which Perplexity mode should you use?<\/a><\/li><li><a href=\"#current-models\">Perplexity AI models: the current Search lineup<\/a><\/li><li><a href=\"#models-vs-modes\">Models, modes, Thinking, and source filters are different<\/a><\/li><li><a href=\"#mode-details\">When to use each mode and setting<\/a><\/li><li><a href=\"#change-model\">How to change the AI model in Perplexity<\/a><\/li><li><a href=\"#best-model\">Which model is best for your task?<\/a><\/li><li><a href=\"#which-plan\">Do you need Free, Pro, or Max?<\/a><\/li><li><a href=\"#api-models\">Are Sonar API models the same as Perplexity app models?<\/a><\/li><li><a href=\"#all-in-one\">When an all-in-one AI workspace makes more sense<\/a><\/li><li><a href=\"#faq\">\u0e04\u0e33\u0e16\u0e32\u0e21\u0e17\u0e35\u0e48\u0e21\u0e31\u0e01\u0e16\u0e39\u0e01\u0e16\u0e32\u0e21<\/a><\/li><li><a href=\"#verdict\">\u0e04\u0e38\u0e13\u0e04\u0e27\u0e23\u0e40\u0e25\u0e37\u0e2d\u0e01\u0e2d\u0e31\u0e19\u0e44\u0e2b\u0e19\u0e14\u0e35?<\/a><\/li><\/ol><\/nav>\n\n\n\n<h2 id=\"which-mode\" class=\"wp-block-heading pplx-heading\">Which Perplexity mode should you use?<\/h2>\n\n\n\n<figure class=\"wp-block-table pplx-table\"><table><thead><tr><th>\u0e07\u0e32\u0e19\u0e02\u0e2d\u0e07\u0e04\u0e38\u0e13<\/th><th>\u0e40\u0e23\u0e34\u0e48\u0e21\u0e15\u0e49\u0e19\u0e14\u0e49\u0e27\u0e22<\/th><th>\u0e17\u0e33\u0e44\u0e21\u0e16\u0e36\u0e07\u0e40\u0e2b\u0e21\u0e32\u0e30\u0e2a\u0e21<\/th><\/tr><\/thead><tbody><tr><td>Find a fact, compare a few products, or catch up on a topic<\/td><td>Search + Best<\/td><td>Get a cited answer without choosing an engine yourself.<\/td><\/tr><tr><td>Work through a complex decision, calculation, or coding question<\/td><td>Search + an eligible model with Thinking<\/td><td>Give the model more reasoning time when the setting is available.<\/td><\/tr><tr><td>Prepare a market overview, literature review, or due-diligence report<\/td><td>\u0e01\u0e32\u0e23\u0e27\u0e34\u0e08\u0e31\u0e22\u0e40\u0e0a\u0e34\u0e07\u0e25\u0e36\u0e01<\/td><td>Use a workflow built to gather and synthesize more evidence.<\/td><\/tr><tr><td>Study a concept and check your understanding<\/td><td>\u0e42\u0e2b\u0e21\u0e14\u0e01\u0e32\u0e23\u0e40\u0e23\u0e35\u0e22\u0e19\u0e23\u0e39\u0e49<\/td><td>Get guided explanations, hints, and questions.<\/td><\/tr><tr><td>Explore disagreement across several AI answers<\/td><td>Model Council<\/td><td>Compare multiple perspectives and read a synthesis.<\/td><\/tr><tr><td>Produce files, build an app, or complete connected work<\/td><td>\u0e04\u0e2d\u0e21\u0e1e\u0e34\u0e27\u0e40\u0e15\u0e2d\u0e23\u0e4c Perplexity<\/td><td>Delegate a task to an agent with tools.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Choose by the output you need. A short answer about a company\u2019s pricing belongs in Search; a competitor landscape with source-backed findings belongs in Deep Research; a spreadsheet and presentation built from those findings belongs in Computer.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">\u0e04\u0e27\u0e32\u0e21\u0e2a\u0e31\u0e1a\u0e2a\u0e19 <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/10354919-what-advanced-ai-models-does-perplexity-pro-unlock\">official model and mode guide<\/a> makes the key distinction explicit: <strong>Best is an automatic model option inside Search, not a separate mode.<\/strong> Thinking is also a setting on a supported model, rather than another model name.<\/p>\n\n\n\n<h2 id=\"current-models\" class=\"wp-block-heading pplx-heading\">Perplexity AI models: the current Search lineup<\/h2>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Perplexity uses both its own models and third-party models. The table below covers the consumer Search model selector, using Perplexity\u2019s September 4, 2026 availability snapshot, checked on September 14. It is not the API catalog, the Computer selector, or a benchmark ranking.<\/p>\n\n\n\n<figure class=\"wp-block-table pplx-table\"><table><thead><tr><th>Search model<\/th><th>\u0e1c\u0e39\u0e49\u0e43\u0e2b\u0e49\u0e1a\u0e23\u0e34\u0e01\u0e32\u0e23<\/th><th>\u0e02\u0e49\u0e2d\u0e14\u0e35<\/th><th>\u0e41\u0e21\u0e47\u0e01\u0e0b\u0e4c<\/th><\/tr><\/thead><tbody><tr><td>Sonar 2<\/td><td>\u0e04\u0e27\u0e32\u0e21\u0e2a\u0e31\u0e1a\u0e2a\u0e19<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>GPT-5.6 Terra<\/td><td>\u0e42\u0e2d\u0e40\u0e1e\u0e48\u0e19\u0e40\u0e2d\u0e44\u0e2d<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>GPT-5.6 Sol<\/td><td>\u0e42\u0e2d\u0e40\u0e1e\u0e48\u0e19\u0e40\u0e2d\u0e44\u0e2d<\/td><td>\u0e44\u0e21\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>Gemini 3.7 Flash<\/td><td>Google<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>\u0e42\u0e04\u0e25\u0e14 \u0e42\u0e0b\u0e40\u0e19\u0e47\u0e15 5<\/td><td>\u0e41\u0e2d\u0e19\u0e42\u0e17\u0e23\u0e1b\u0e34\u0e01<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>Claude Opus 5<\/td><td>\u0e41\u0e2d\u0e19\u0e42\u0e17\u0e23\u0e1b\u0e34\u0e01<\/td><td>\u0e44\u0e21\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>Kimi K3<\/td><td>Moonshot AI<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>GLM 5.3<\/td><td>Z.ai<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>Grok 4.6<\/td><td>\u0e40\u0e2d\u0e47\u0e01\u0e0b\u0e4c\u0e44\u0e2d\u0e44\u0e2d<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><tr><td>Nemotron 3 Ultra<\/td><td>NVIDIA<\/td><td>\u0e43\u0e0a\u0e48<\/td><td>\u0e43\u0e0a\u0e48<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">\u0e41\u0e2b\u0e25\u0e48\u0e07\u0e17\u0e35\u0e48\u0e21\u0e32: <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/10354919-what-advanced-ai-models-does-perplexity-pro-unlock\">Perplexity\u2019s subscription model availability table<\/a>. Free users can use Best and limited Pro Searches, but cannot manually choose an advanced Search model. Perplexity says the account\u2019s model selector is the source of truth because rollouts, promotions, regional requirements, and organization settings can affect access.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">The Enterprise table is different: it lists all ten of these models for both Enterprise Pro and Enterprise Max, subject to administrator and organization controls. Do not apply the consumer Pro restrictions to an Enterprise Pro account.<\/p>\n\n\n\n<p class=\"pplx-note wp-block-paragraph\"><strong>Pro or Max?<\/strong> For consumer Search access, GPT-5.6 Sol and Claude Opus 5 are the two Max-only entries in this snapshot. That is a reason to consider Max if you need those specific models; it does not make every ordinary search worth upgrading for.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">OpenAI models inside Perplexity are not the ChatGPT application. Our <a href=\"https:\/\/www.glbgpt.com\/hub\/does-perplexity-use-chatgpt-the-truth-you-need-to-know\/\">explanation of Perplexity\u2019s connection to ChatGPT<\/a> covers that relationship; model access, app features, subscriptions, and saved conversations remain separate concepts.<\/p>\n\n\n\n<h2 id=\"models-vs-modes\" class=\"wp-block-heading pplx-heading\">Models, modes, Thinking, and source filters are different<\/h2>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">A <strong>\u0e41\u0e1a\u0e1a\u0e08\u0e33\u0e25\u0e2d\u0e07<\/strong> is the engine producing an answer. A <strong>\u0e42\u0e2b\u0e21\u0e14<\/strong> organizes the work. <strong>\u0e01\u0e32\u0e23\u0e04\u0e34\u0e14<\/strong> gives a supported engine more time to reason. Your <strong>\u0e41\u0e1c\u0e19<\/strong> determines the choices and allowances you can use. These controls can interact, but changing one does not automatically change the others.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full pplx-figure\"><a href=\"https:\/\/static.futureshareai.com\/glb_features\/perplexity-model-mode-thinking-plan_f045afa5bf8d4ce3beb5ca5e2a7cf7ec.webp\"><img decoding=\"async\" src=\"https:\/\/static.futureshareai.com\/glb_features\/perplexity-model-mode-thinking-plan_f045afa5bf8d4ce3beb5ca5e2a7cf7ec.webp\" alt=\"Four separate Perplexity controls: mode for workflow, model for the answer engine, Thinking for reasoning, and plan for access\"\/><\/a><figcaption class=\"wp-element-caption\">Mode, model, Thinking, and plan answer four different questions.<\/figcaption><\/figure>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Source selection is another distinction. Choosing where a search should look is different from choosing which model writes the answer. A research task can fail because the sources are unsuitable even when the underlying model is capable.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Older tutorials often group source choices under Focus. Our <a href=\"https:\/\/www.glbgpt.com\/hub\/what-are-the-different-focus-modes-in-perplexity-ai-full-guide-2025\/\">overview of Perplexity Focus and source selection<\/a> provides background on that terminology. For the current model names and workflow labels, use the dated Search table above and the options in your own account.<\/p>\n\n\n\n<h2 id=\"mode-details\" class=\"wp-block-heading pplx-heading\">When to use each mode and setting<\/h2>\n\n\n\n<figure class=\"wp-block-image size-full pplx-figure\"><a href=\"https:\/\/static.futureshareai.com\/glb_features\/01-perplexity-search-modes_52abd6288c7245bb905e4d93ef7e4d7e.webp\"><img decoding=\"async\" src=\"https:\/\/static.futureshareai.com\/glb_features\/01-perplexity-search-modes_52abd6288c7245bb905e4d93ef7e4d7e.webp\" alt=\"Perplexity Search menu showing Search, Deep research, Model council, and Learn step by step\"\/><\/a><figcaption class=\"wp-element-caption\">The Search menu separates everyday search, Deep research, Model council, and guided learning. Locked options are shown in this signed-out view. <a href=\"https:\/\/www.perplexity.ai\/\">Source: Perplexity<\/a><\/figcaption><\/figure>\n\n\n\n<h3 id=\"search-best\" class=\"wp-block-heading pplx-heading\">Search + Best: the everyday default<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Start here for definitions, quick checks, current information, and small comparisons. Best chooses a suitable model for the query. You avoid spending more time selecting an engine than asking the question.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">If an answer is weak, first make the question more specific: add the date, country, product version, or sources you need. A vague prompt can produce a vague answer across several models. Open the citations before acting on a claim.<\/p>\n\n\n\n<h3 id=\"thinking\" class=\"wp-block-heading pplx-heading\">Thinking: difficult reasoning in an answer<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Use Thinking when a question has several dependencies: debugging a function, weighing conflicting requirements, planning a migration, or checking a calculation. It can help the model reason more carefully, at the cost of extra waiting time. Availability depends on the selected model; some reasoning-focused models use it automatically.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">A useful request is: \u201cCheck the assumptions, work through the constraints, and explain the result so I can verify it.\u201d Extra reasoning time does not guarantee correct facts or working code.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">For programming work, our <a href=\"https:\/\/www.glbgpt.com\/hub\/is-perplexity-good-for-coding\/\">guide to using Perplexity for coding<\/a> separates documentation research, code explanation, and debugging from a full development workflow. Run generated code and tests in your own project before relying on it.<\/p>\n\n\n\n<h3 id=\"deep-research\" class=\"wp-block-heading pplx-heading\">Deep Research: a report assembled from evidence<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Use Deep Research when the job needs breadth and synthesis: a market analysis, literature review, or comparison of competing explanations. Perplexity selects the models automatically, so manually choosing a Search model is not how you control this workflow.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">\u0e04\u0e27\u0e32\u0e21\u0e2a\u0e31\u0e1a\u0e2a\u0e19 <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/13600190-what-s-new-in-advanced-deep-research\">Advanced Deep Research documentation<\/a> describes broader source gathering, cross-referencing, uploaded-document support, calculations, and a code sandbox. Give it a clear scope and report format so those capabilities serve a defined question.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Research workflows also differ between products. The <a href=\"https:\/\/www.glbgpt.com\/hub\/how-to-use-chatgpt-deep-research-complete-tutorial-tips-and-best-practices\/\">\u0e04\u0e39\u0e48\u0e21\u0e37\u0e2d\u0e01\u0e32\u0e23\u0e27\u0e34\u0e08\u0e31\u0e22\u0e40\u0e0a\u0e34\u0e07\u0e25\u0e36\u0e01 ChatGPT<\/a> is useful if you are deciding where to carry out a longer investigation, rather than simply which Search model to select.<\/p>\n\n\n\n<h3 id=\"learn-mode\" class=\"wp-block-heading pplx-heading\">Learn Mode: understanding before an answer<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Choose Learn Mode for guided explanations, hints, questions, and quizzes. It suits a student who wants to understand why a solution works or a professional learning a new subject.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">The <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/12120542-what-is-learn-mode\">Learn Mode help page<\/a> describes availability for all users but says the visible toggle is limited to verified students. If you cannot see it, do not assume your model is missing: the access route and the selected model are separate issues.<\/p>\n\n\n\n<h3 id=\"model-council\" class=\"wp-block-heading pplx-heading\">Model Council: compare perspectives<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Model Council runs a question across several models and synthesizes their answers. Use it when disagreement matters: identifying the weakest assumption in a proposal, reviewing a decision from different angles, or finding claims that deserve another source check.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">The <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/13641704-what-is-model-council\">dedicated Model Council documentation<\/a> distinguishes two routes. On web and iOS, it is included with Max and Enterprise Max and uses three models. In Computer, it runs as a skill on all paid plans, supports two to eight models, and uses Computer credits.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full pplx-figure\"><a href=\"https:\/\/static.futureshareai.com\/glb_features\/02-perplexity-model-council-result_43b6ecb1e4534306a150af57a5288025.webp\"><img decoding=\"async\" src=\"https:\/\/static.futureshareai.com\/glb_features\/02-perplexity-model-council-result_43b6ecb1e4534306a150af57a5288025.webp\" alt=\"Official Perplexity Model Council example showing a Where Models Agree table and supporting evidence\"\/><\/a><figcaption class=\"wp-element-caption\">Perplexity\u2019s official Model Council example groups shared findings under \u201cWhere Models Agree\u201d alongside evidence. The example\u2019s \u201c2 models\u201d label differs from the help page\u2019s three-model web instructions. <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/13641704-what-is-model-council\">Source: Perplexity<\/a><\/figcaption><\/figure>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Agreement is useful, but it is not proof. Several models can repeat the same weak source. Read the disagreements and verify the underlying evidence before making a consequential decision.<\/p>\n\n\n\n<h3 id=\"computer-comet\" class=\"wp-block-heading pplx-heading\">Computer and Comet: when the task goes beyond an answer<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\"><a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/13837784-what-is-computer\">\u0e04\u0e2d\u0e21\u0e1e\u0e34\u0e27\u0e40\u0e15\u0e2d\u0e23\u0e4c Perplexity<\/a> is for work that requires tools and deliverables: researching competitors, building a spreadsheet, creating a presentation, or working with connected apps. It requires an active subscription and available credits; its model selector and usage rules are separate from Search.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Comet is Perplexity\u2019s browser environment for browsing assistance and browser tasks. It also has its own model-selection experience. Choose the product because you need browser help, rather than treating Comet as another model in the Search table.<\/p>\n\n\n\n<h2 id=\"change-model\" class=\"wp-block-heading pplx-heading\">How to change the AI model in Perplexity<\/h2>\n\n\n\n<figure class=\"wp-block-image size-full pplx-figure\"><a href=\"https:\/\/static.futureshareai.com\/glb_features\/03-perplexity-search-model-selector_54f1d835d4ed457f9bd37f7d481978e2.webp\"><img decoding=\"async\" src=\"https:\/\/static.futureshareai.com\/glb_features\/03-perplexity-search-model-selector_54f1d835d4ed457f9bd37f7d481978e2.webp\" alt=\"Perplexity Model menu on the right of the Search input, showing advanced models with lock icons in a signed-out session\"\/><\/a><figcaption class=\"wp-element-caption\">Open Model on the right of the Search input. This signed-out view lists advanced models with locks; usable options depend on your account and plan. <a href=\"https:\/\/www.perplexity.ai\/\">Source: Perplexity<\/a><\/figcaption><\/figure>\n\n\n\n<ol class=\"wp-block-list pplx-list\">\n<li>Open Perplexity and start a Search query. Use an account with a paid plan if you want manual advanced-model selection.<\/li>\n\n\n\n<li>Open the Model control in the prompt input box. In the web interface captured on September 14, 2026, it is on the right; the Search mode menu is on the left.<\/li>\n\n\n\n<li>Choose an eligible model, or leave the selection on Best. Turn on Thinking if the chosen model offers it and the task needs more reasoning.<\/li>\n\n\n\n<li>Submit a specific prompt, inspect the answer and citations, and change one setting at a time if you want to compare the result.<\/li>\n<\/ol>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">These steps follow Perplexity\u2019s <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/12802813-how-do-i-change-to-my-preferred-ai-model\">official model-switching instructions<\/a>. The location is documented for the prompt input; the exact options depend on your account and workflow.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">If you are new to the product, start with the <a href=\"https:\/\/www.glbgpt.com\/hub\/how-to-use-perplexity-ai-a-complete-beginners-guide\/\">Perplexity beginner\u2019s guide<\/a> for asking questions and reading citations. Use the current selector for model and mode names, since older interface labels can persist in tutorials.<\/p>\n\n\n\n<figure class=\"wp-block-table pplx-table\"><table><thead><tr><th>What you notice<\/th><th>\u0e04\u0e33\u0e2d\u0e18\u0e34\u0e1a\u0e32\u0e22\u0e17\u0e35\u0e48\u0e40\u0e1b\u0e47\u0e19\u0e44\u0e1b\u0e44\u0e14\u0e49<\/th><th>\u0e2a\u0e34\u0e48\u0e07\u0e17\u0e35\u0e48\u0e04\u0e27\u0e23\u0e17\u0e33<\/th><\/tr><\/thead><tbody><tr><td>Only Best is available<\/td><td>Free access or a restricted account<\/td><td>Check the plan and account you are signed into.<\/td><\/tr><tr><td>A model in this article is missing<\/td><td>Different plan, rollout, region, or Enterprise policy<\/td><td>Compare your plan with the current official list; ask your organization administrator if needed.<\/td><\/tr><tr><td>No manual model picker in Deep Research<\/td><td>The workflow chooses models automatically<\/td><td>Specify scope, sources, and output instead.<\/td><\/tr><tr><td>Thinking is not offered<\/td><td>The selected model does not expose that setting, or uses reasoning automatically<\/td><td>Inspect that model\u2019s available options.<\/td><\/tr><tr><td>The same model gives different answers elsewhere<\/td><td>Different app tools, retrieved sources, instructions, or context<\/td><td>Compare the complete workflow and evidence, not just the model label.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<h2 id=\"best-model\" class=\"wp-block-heading pplx-heading\">Which model is best for your task?<\/h2>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\"><strong>There is no verified universal winner in the access table.<\/strong> It tells you what is available, not which model will do your work best. Start with Best, then compare eligible models on a real task only when the default falls short.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Keep the prompt, attachments, and source requirements unchanged. Judge whether the answer is correct, whether its citations support the key claims, and how much editing you need. Prefer the result that helps you finish the job over the one that sounds most confident.<\/p>\n\n\n\n<h3 id=\"research-prompt\" class=\"wp-block-heading pplx-heading\">For research: define the evidence standard<\/h3>\n\n\n\n<pre class=\"wp-block-code pplx-code\"><code>Compare three project-management tools for a 12-person design agency. Use official pricing and product documentation, dated as of today. Separate confirmed features from unclear claims. Return a comparison table, five decision criteria, and a short recommendation with citations.<\/code><\/pre>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Use Search for an initial shortlist. Choose Deep Research if the decision needs a longer investigation across integrations, constraints, and conflicting evidence. Add your real budget and must-have features before running the prompt.<\/p>\n\n\n\n<h3 id=\"coding-prompt\" class=\"wp-block-heading pplx-heading\">For coding: make the answer verifiable<\/h3>\n\n\n\n<pre class=\"wp-block-code pplx-code\"><code>Explain why this function fails on an empty list. Propose the smallest fix, preserve the public interface, and include tests for empty input, one item, and repeated values. Cite official documentation for any library behavior you rely on.\n\n&#91;Paste the function and exact error here.]<\/code><\/pre>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Use a supported model with Thinking for a problem that needs multi-step analysis. Compare proposed fixes by running them, not by counting how many paragraphs the model produces.<\/p>\n\n\n\n<h3 id=\"study-prompt\" class=\"wp-block-heading pplx-heading\">For studying: request active recall<\/h3>\n\n\n\n<pre class=\"wp-block-code pplx-code\"><code>Help me understand Bayes\u2019 theorem. First ask what I already know. Use one everyday example, give hints before the solution, and finish with three short questions that test whether I can apply it.<\/code><\/pre>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Use Learn Mode when available. Adapt the prompt to your subject and level, then check whether you can answer the follow-up questions without help.<\/p>\n\n\n\n<h2 id=\"which-plan\" class=\"wp-block-heading pplx-heading\">Do you need Free, Pro, or Max?<\/h2>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Stay on Free if basic answers and limited Pro Searches meet your needs. Consider Pro when you repeatedly need manual Search model selection. Consider Max for specific Max-only Search models, the web\/iOS version of Model Council, or higher usage allowances.<\/p>\n\n\n\n<figure class=\"wp-block-table pplx-table\"><table><thead><tr><th>\u0e41\u0e1e\u0e47\u0e01\u0e40\u0e01\u0e08\u0e2a\u0e33\u0e2b\u0e23\u0e31\u0e1a\u0e1c\u0e39\u0e49\u0e1a\u0e23\u0e34\u0e42\u0e20\u0e04<\/th><th>Verified public pricing basis<\/th><th>\u0e40\u0e25\u0e37\u0e2d\u0e01\u0e43\u0e0a\u0e49\u0e40\u0e21\u0e37\u0e48\u0e2d<\/th><\/tr><\/thead><tbody><tr><td>\u0e1f\u0e23\u0e35<\/td><td>$0<\/td><td>You can work with Best and the free allowances.<\/td><\/tr><tr><td>\u0e02\u0e49\u0e2d\u0e14\u0e35<\/td><td>The official product page displays $17\/month when billed annually.<\/td><td>You want manual access to the eligible Pro Search models.<\/td><\/tr><tr><td>\u0e41\u0e21\u0e47\u0e01\u0e0b\u0e4c<\/td><td>$200\/month or $2,000\/year; the product page rounds the annual equivalent to $167\/month.<\/td><td>You need Max-only Search models, web\/iOS Model Council, or more capacity.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Pricing checked September 14, 2026. The annual-equivalent Pro and Max display comes from <a href=\"https:\/\/www.perplexity.ai\/pro\">Perplexity\u2019s official subscription page<\/a>. \u201cPer month when billed annually\u201d is not a month-to-month charge; confirm the full annual total and applicable taxes at checkout.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">The exact Max monthly and yearly amounts are stated in the <a href=\"https:\/\/www.perplexity.ai\/help-center\/en\/articles\/13641704-what-is-model-council\">Model Council pricing explanation<\/a>. For Pro, the price above is the annual-billing equivalent; check checkout for the month-to-month offer.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">For a broader feature comparison, see <a href=\"https:\/\/www.glbgpt.com\/hub\/perplexity-pro-vs-max-features-2026\/\">Perplexity Pro versus Max<\/a>. Match any purchase to the current checkout and usage allowances, rather than assuming that paid access means unlimited use of every model and workflow.<\/p>\n\n\n\n<h2 id=\"api-models\" class=\"wp-block-heading pplx-heading\">Are Sonar API models the same as Perplexity app models?<\/h2>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">No. API models are selected programmatically and billed separately from the consumer app. A Pro or Max subscription does not make API usage part of the same allowance. The Search table above is not a list of API model IDs.<\/p>\n\n\n\n<figure class=\"wp-block-table pplx-table\"><table><thead><tr><th>Legacy Sonar API model<\/th><th>Documented purpose<\/th><\/tr><\/thead><tbody><tr><td>sonar<\/td><td>Lightweight web-grounded answers.<\/td><\/tr><tr><td>sonar-pro<\/td><td>More advanced search and follow-up questions.<\/td><\/tr><tr><td>sonar-reasoning-pro<\/td><td>Multi-step reasoning with web grounding.<\/td><\/tr><tr><td>sonar-deep-research<\/td><td>Extended research and report synthesis.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">The <a href=\"https:\/\/docs.perplexity.ai\/docs\/sonar\/models\">Sonar API model documentation<\/a> describes this legacy family. These names should not be treated as interchangeable with Sonar 2 in the consumer Search selector.<\/p>\n\n\n\n<p class=\"pplx-note wp-block-paragraph\"><strong>Developer update:<\/strong> Perplexity says Sonar Chat Completions will be supported until <strong>September 27, 2026<\/strong> and recommends the Agent API for new projects. Existing integrations should follow the <a href=\"https:\/\/docs.perplexity.ai\/docs\/agent-api\/migrate-from-sonar\/overview\">official Sonar-to-Agent API migration guide<\/a>. This API change is not a notice that the Perplexity consumer app is shutting down.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">For a new integration, use the <a href=\"https:\/\/docs.perplexity.ai\/docs\/agent-api\/models\">Agent API model catalog<\/a> to select supported model identifiers and check token pricing. Keep that developer decision separate from choosing Search, Deep Research, or a consumer subscription.<\/p>\n\n\n\n<h2 id=\"all-in-one\" class=\"wp-block-heading pplx-heading\">When an all-in-one AI workspace makes more sense<\/h2>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Perplexity is a strong fit when the main job is finding information and producing source-backed answers. A broader workspace makes sense when research is only the first step and you also need drafting, image creation, or video production.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\"><a href=\"https:\/\/www.glbgpt.com\/home?inviter=hub_popup&amp;login=1\">\u0e42\u0e01\u0e25\u0e1a\u0e2d\u0e25\u0e08\u0e35\u0e1e\u0e35\u0e17\u0e35<\/a> brings multiple supported model families into one account and subscription. Its practical benefit is consolidating access and reducing repeated logins across provider websites. Choose it for that broader workflow; do not assume it reproduces Perplexity\u2019s native Model Council, Computer, or account controls.<\/p>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">If your main requirement is still unclear, compare the <a href=\"https:\/\/www.glbgpt.com\/hub\/11-best-perplexity-ai-alternatives-in-2026\/\">Perplexity alternatives grouped by workflow<\/a> before buying another subscription. A specialist research tool and an all-in-one creative workspace solve different combinations of tasks.<\/p>\n\n\n\n<aside class=\"pplx-cta\"><p><strong>Research is one part of your workflow.<\/strong><br>Bring writing, images, and video into one AI workspace when you need more than search.<\/p><a href=\"https:\/\/www.glbgpt.com\/home?inviter=hub_popup&amp;login=1\">\u0e04\u0e49\u0e19\u0e1e\u0e1a GlobalGPT<\/a><\/aside>\n\n\n\n<h2 id=\"faq\" class=\"wp-block-heading pplx-heading\">\u0e04\u0e33\u0e16\u0e32\u0e21\u0e17\u0e35\u0e48\u0e21\u0e31\u0e01\u0e16\u0e39\u0e01\u0e16\u0e32\u0e21<\/h2>\n\n\n\n<h3 id=\"faq-1\" class=\"wp-block-heading pplx-heading\">What AI models does Perplexity use?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Perplexity combines its own Sonar models with third-party models. Its September 4, 2026 Search list includes OpenAI, Anthropic, Google, Moonshot AI, Z.ai, xAI, and NVIDIA options. Access depends on your plan and account.<\/p>\n\n\n\n<h3 id=\"faq-2\" class=\"wp-block-heading pplx-heading\">What is the best Perplexity model for most users?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Start with Best inside Search. It automatically picks a suitable model for your question. Use Thinking for complex reasoning or Deep Research for a full report before assuming you need a different engine.<\/p>\n\n\n\n<h3 id=\"faq-3\" class=\"wp-block-heading pplx-heading\">Can I choose an AI model on the free plan?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Free users can use Best and limited Pro Searches, but cannot manually select an advanced Search model. Manual selection is available on paid plans, subject to model availability and usage limits.<\/p>\n\n\n\n<h3 id=\"faq-4\" class=\"wp-block-heading pplx-heading\">Is Best a separate Perplexity mode?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">No. Best is the automatic model option inside Search. Search is the workflow; Best lets Perplexity choose an appropriate model for each query.<\/p>\n\n\n\n<h3 id=\"faq-5\" class=\"wp-block-heading pplx-heading\">Can I choose the model used by Deep Research?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">Perplexity selects Deep Research models automatically. Focus on your research question, source requirements, attachments, and output format rather than assuming the Search model picker controls the report.<\/p>\n\n\n\n<h3 id=\"faq-6\" class=\"wp-block-heading pplx-heading\">Is Model Council only available with Max?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">On web and iOS, Model Council requires Max or Enterprise Max. In Perplexity Computer, it is available as a skill on all paid plans and uses Computer credits. The two routes have different limits and model counts.<\/p>\n\n\n\n<h3 id=\"faq-7\" class=\"wp-block-heading pplx-heading\">Does Perplexity Pro include ChatGPT Plus or Claude Pro?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">No. Access to an OpenAI or Anthropic model inside Perplexity is not a subscription to the provider\u2019s own app. Features, saved conversations, account controls, and usage allowances belong to the product you use.<\/p>\n\n\n\n<h3 id=\"faq-8\" class=\"wp-block-heading pplx-heading\">Is GlobalGPT the same as Perplexity?<\/h3>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\">No. GlobalGPT is an all-in-one platform for multiple supported AI models and creative workflows. Perplexity has its own native search, research, and agent features. Choose based on the work you need to complete.<\/p>\n\n\n\n<h2 id=\"verdict\" class=\"wp-block-heading pplx-heading\">\u0e04\u0e38\u0e13\u0e04\u0e27\u0e23\u0e40\u0e25\u0e37\u0e2d\u0e01\u0e2d\u0e31\u0e19\u0e44\u0e2b\u0e19\u0e14\u0e35?<\/h2>\n\n\n\n<p class=\"pplx-text wp-block-paragraph\"><strong>Start with Search + Best.<\/strong> Add Thinking for a difficult reasoning task. Move to Deep Research for a substantial report, Learn Mode for studying, Model Council for multiple perspectives, or Computer for completed deliverables. Upgrade only when a specific model, feature, or usage limit is holding your work back.<\/p>\n\n\n\n<script type=\"application\/ld+json\">{\n    \"@context\": \"https:\\\/\\\/schema.org\",\n    \"@type\": \"FAQPage\",\n    \"mainEntity\": [\n        {\n            \"@type\": \"Question\",\n            \"name\": \"What AI models does Perplexity use?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Perplexity combines its own Sonar models with third-party models. Its September 4, 2026 Search list includes OpenAI, Anthropic, Google, Moonshot AI, Z.ai, xAI, and NVIDIA options. Access depends on your plan and account.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"What is the best Perplexity model for most users?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Start with Best inside Search. It automatically picks a suitable model for your question. Use Thinking for complex reasoning or Deep Research for a full report before assuming you need a different engine.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Can I choose an AI model on the free plan?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Free users can use Best and limited Pro Searches, but cannot manually select an advanced Search model. Manual selection is available on paid plans, subject to model availability and usage limits.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Is Best a separate Perplexity mode?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"No. Best is the automatic model option inside Search. Search is the workflow; Best lets Perplexity choose an appropriate model for each query.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Can I choose the model used by Deep Research?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Perplexity selects Deep Research models automatically. Focus on your research question, source requirements, attachments, and output format rather than assuming the Search model picker controls the report.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Is Model Council only available with Max?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"On web and iOS, Model Council requires Max or Enterprise Max. In Perplexity Computer, it is available as a skill on all paid plans and uses Computer credits. The two routes have different limits and model counts.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Does Perplexity Pro include ChatGPT Plus or Claude Pro?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"No. Access to an OpenAI or Anthropic model inside Perplexity is not a subscription to the provider\\u2019s own app. Features, saved conversations, account controls, and usage allowances belong to the product you use.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Is GlobalGPT the same as Perplexity?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"No. GlobalGPT is an all-in-one platform for multiple supported AI models and creative workflows. Perplexity has its own native search, research, and agent features. Choose based on the work you need to complete.\"\n            }\n        }\n    ]\n}<\/script>","protected":false},"excerpt":{"rendered":"<p>For most people, the best starting point is Search with Best selected. Perplexity AI models are the engines behind its answers; modes decide whether you get a quick answer, a research report, a lesson, or completed work. Use Thinking for a difficult multi-step question and Deep Research when you need a substantial report with sources. [&hellip;]<\/p>","protected":false},"author":17,"featured_media":19347,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_seopress_robots_primary_cat":"","_seopress_titles_title":"Perplexity AI Models and Modes: Which One Should You Use?","_seopress_titles_desc":"Compare Perplexity AI models and modes with the current access table. See when to use Best, Thinking, or Deep Research and choose the right setup.","_seopress_robots_index":"","footnotes":""},"categories":[7],"tags":[],"class_list":["post-19346","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-chat"],"_links":{"self":[{"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/posts\/19346","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/users\/17"}],"replies":[{"embeddable":true,"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/comments?post=19346"}],"version-history":[{"count":1,"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/posts\/19346\/revisions"}],"predecessor-version":[{"id":19348,"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/posts\/19346\/revisions\/19348"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/media\/19347"}],"wp:attachment":[{"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/media?parent=19346"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/categories?post=19346"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/wp.glbgpt.com\/th\/wp-json\/wp\/v2\/tags?post=19346"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}