Move your team from Perplexity to ChatGPT.
Each person gets a short interview, then a personalized step-by-step guide showing exactly where their saved work, prompts, and projects land in ChatGPT — and the few things that need a manual step, walked through. You send one invite; nobody becomes the help desk, and nothing in Perplexity is changed or cancelled. Want to see where you’re paying for both first? Run the free audit.
ChatGPT lets users build named, sharable Custom GPTs via the no-code GPT Builder at chatgpt.com/gpts, with a public GPT Store and optional Actions (API calls). Perplexity has no GPT-Builder analog. The closest pattern is to configure a Space with custom instructions, pinned files, and a chosen default model - that Space then behaves like a reusable assistant for its members. There is no marketplace, no discrete 'Assistant' object, and no Actions equivalent.
- Warning: Asymmetric: Perplexity cannot replicate GPT Actions (custom API tool-calling) at all. Use Perplexity connectors (Gmail/GitHub/etc.) instead, but they are a fixed list, not user-defined.
- Warning: No public discoverability: a Perplexity Space can only be shared via link to specific users; there is no Store.
- Warning: Sharing a Custom GPT preserves its instructions for any chatgpt.com user with the link; sharing a Space requires the recipient to also be a Perplexity user.
Both products ship a long-running multi-source research agent that produces a cited structured report. ChatGPT Deep Research (Feb 2026 update, GPT-5.2-based) runs 5-30 minutes, can pull from connected Apps via MCP, and is gated as: Free ~5/mo, Plus/Team/Enterprise/Edu 10/mo (reduced from 25 on Feb 25 2026), Pro highest (the $100 tier ~5x Plus, the $200 tier ~250/mo; overflow drops to a lightweight model). Perplexity Deep Research runs 2-4 minutes; after the Feb-2026 reduction Perplexity Pro is capped at roughly 20/month (Free gets only a small allowance). Output formats and citation style are similar; Perplexity also offers Labs for longer agentic artifact-building (~10+ minutes, builds spreadsheets/dashboards/mini-apps, ~25/month on Pro).
- Warning: Volume flipped in 2026: ChatGPT Pro now offers far more Deep Research headroom (250/mo) than Perplexity Pro (~20/mo after the Feb-2026 cut) - the opposite of the older assumption that Perplexity had the higher caps.
- Warning: Perplexity Labs has no direct ChatGPT analog - the closest is ChatGPT Agent (browser/computer-use agent) but Agent is action-oriented, not artifact-building.
- Warning: ChatGPT Deep Research can restrict scope to selected sites and use MCP-connected tools; Perplexity Deep Research uses its own crawler/index and the configured connectors but no MCP.
Both products accept PDFs, Office docs, text, code, images, and (consumer) audio/video via the prompt bar. ChatGPT caps files at 512 MB hard / ~25 MB recommended per file, 2M tokens for text, with ~80 files per 3h on Plus and unlimited on Pro; Free is ~3/day. Perplexity caps consumer files at ~25-40 MB per file, up to 10 per message; Free is ~3/day, Pro is unlimited per-thread plus up to 50 persisted in a Space. For long-running knowledge bases, prefer ChatGPT Projects or Perplexity Spaces over per-thread uploads.
- Warning: Perplexity per-file size limit is inconsistent across sources (25 MB vs 40 MB consumer, 50 MB Enterprise) - re-verify against the help center for the specific plan.
- Warning: ChatGPT's 2M-token text cap can reject large PDFs even when they're under 512 MB. There is no equivalent token cap published for Perplexity.
- Warning: Files attached only to a single ChatGPT chat are not pinned to a Project automatically; same in Perplexity - thread uploads are temporary and auto-delete after 7 days unless added to a Space.
ChatGPT Memory (Settings > Personalization > Memory) has two layers: explicit 'saved memories' (editable list) and implicit 'reference chat history' (recall from past chats, no list). Perplexity's analog is the Library (perplexity.ai/library, every thread sorted/searchable) plus an opt-in Memory layer that synthesizes durable preferences and cites the memory item when it surfaces. Both can be turned off entirely, and both exclude Incognito/temporary chats.
- Warning: ChatGPT's reference-chat-history layer has no list view - users see effects but can't audit it directly. Perplexity surfaces memory items with citations, which is more transparent.
- Warning: Clearing memory in either product does not delete the underlying conversation history; you must also delete chats from the Library/history list separately.
- Warning: Migration is one-way and manual: there's no export of 'saved memories' from ChatGPT, so users have to copy/paste durable facts into Perplexity's Profile or memory layer.
ChatGPT's persistent personalization lives in Settings > Personalization (Custom Instructions text fields, ~1500 char cap, plus a 'base style and tone' preset). Perplexity's analog is Settings > Profile (perplexity.ai/account): a free-text 'About me', preferred language, location, and formatting/tone preferences applied to every non-Incognito thread. Both apply globally; if migrating, copy the 'About me' + 'How should ChatGPT respond' fields straight into Perplexity's Profile text box. Both are available on free plans.
- Warning: Perplexity Profile fields are single text blobs - no separate 'about you' vs 'how to respond' split like ChatGPT.
- Warning: Perplexity's Profile does NOT include a memory layer; that lives in a separate Memory toggle (see memory pattern).
- Warning: Neither product applies these instructions inside Incognito / temporary chats.
ChatGPT SKUs as of July 2026: Free ($0), Go ($8/mo, US, ad-supported), Plus ($20/mo, GPT-5.5 + Deep Research + Agent), Pro ($100/mo since Apr 9 2026, plus higher-tier Pro at $200/mo), Business ($20/seat/mo annual or $25/seat/mo monthly, dropped from $25/$30 on Apr 2 2026; 2-seat minimum), Enterprise (custom, ~$60-100+/seat). Perplexity SKUs: Free, Pro $20/mo ($200/yr), Max $200/mo ($2,000/yr), Education Pro $10/mo (SheerID-verified), Enterprise Pro $40/seat/mo ($400/yr; $30/seat edu discount), Enterprise Max $325/seat/mo ($3,250/yr). The $20/mo individual tier maps cleanly; the $200/mo Max tier maps cleanly to ChatGPT Pro $200. There is no Perplexity equivalent of the $100 ChatGPT Pro intermediate, and Perplexity has no Go/ad-supported tier.
- Warning: Perplexity Max ($200) unlocks Claude Opus 4.5/4.6 and Comet model selection; ChatGPT Pro ($200) unlocks 250/mo Deep Research and unlimited everything but doesn't expose competitor models.
- Warning: Perplexity Enterprise Max ($325/seat) is the high-end - has no direct ChatGPT analog beyond custom-negotiated Enterprise contracts.
- Warning: ChatGPT Business pricing dropped to ~$25/seat in April 2026; older docs may quote $30. Verify before quoting numbers.
- Warning: Perplexity has Education Pro $10/mo (student-verified); ChatGPT Edu is sold per-institution, not per-student - no clean consumer education tier.
- Warning: Price held but value fell on Perplexity's $20 Pro tier: Deep Research and Labs quotas were cut sharply in Feb 2026 (~20 Deep Research/mo, ~25 Labs/mo) - weigh this in a price-for-value comparison.
ChatGPT Projects (sidebar > Projects, gear icon for project-scoped instructions and files) and Perplexity Spaces (left-nav > Spaces) map almost one-to-one: both group threads, attached files, and project-scoped custom instructions for an ongoing topic. To migrate, recreate the project in the target with the same instruction text, then re-upload reference files. Both override global custom instructions inside their scope.
- Warning: ChatGPT Projects are available on all plans including Free (with smaller per-project file caps, ~5 files vs 25 on Plus/40 on Pro). Perplexity Spaces are likewise available on Free with smaller file caps.
- Warning: File caps differ: ChatGPT Team Projects allow up to 40 files; Perplexity Pro Spaces hold up to 50, Enterprise Pro 500, Enterprise Max 5,000.
- Warning: Perplexity Spaces support multi-user collaborators (viewer/collaborator roles). ChatGPT Projects can now be shared with teammates too (members can view chats/files and add their own), so both support in-project collaboration.
Both products ship a per-thread 'Share' button that mints a public, view-only URL. ChatGPT manages shared links at Settings > Data Controls > Shared Links; OpenAI removed the search-engine indexing option after the 2025 incident, but already-indexed snapshots persist. Perplexity has the same model (Share button on a thread copies a public link), and additionally supports Pages as a higher-effort, publishable artifact format.
- Warning: Shared ChatGPT chats from before mid-2025 may still be indexed by Google even though the indexing toggle was removed; advise users to audit and delete old shared links during migration.
- Warning: Both share-link formats are static snapshots - viewer continuing the conversation creates a fork, not edits in place.
- Warning: Perplexity offers an additional layer (Pages) that has no clean ChatGPT analog - it is more like a CMS post than a chat share.
ChatGPT calls them 'Apps' (renamed from Connectors in Dec 2025) and ships ~15 including Google Drive, Gmail, Google Calendar, Outlook, SharePoint, Dropbox, Box, GitHub, Linear, HubSpot, Teams; managed at Settings > Apps. Perplexity offers a similar core list (Gmail, Gcal, Notion, GitHub, Linear, Outlook, Google Drive, OneDrive, SharePoint, Dropbox, Box, Plaid) at perplexity.ai/account/connectors. Both products surface results with citations and let connectors feed into Deep Research. Pair migration: re-OAuth each connector on the destination; permission scopes will differ.
- Warning: Both products restrict connectors by plan: ChatGPT requires a paid plan and excludes EEA/UK/CH for many connectors; Perplexity gates file-store connectors (Drive, SharePoint, OneDrive, Dropbox, Box) to Enterprise Pro/Max.
- Warning: Perplexity has Plaid for personal-finance brokerage sync; ChatGPT has nothing equivalent.
- Warning: Action depth differs: Perplexity's Gmail/Gcal connectors can send mail and create invites; ChatGPT Apps are mostly read/search plus a smaller set of write actions per app.
- Warning: ChatGPT calls connectors 'Apps' since Dec 17, 2025 - migration docs may still use old terminology.
Web search in ChatGPT is one tool among many: a toggleable 'Search' affordance in the prompt bar that returns cited links inline; the model decides when to issue queries. In Perplexity, search IS the product: every thread runs a search by default, with explicit modes (default Search, Pro Search for multi-step, Deep Research for long-running, Labs for agentic artifact-building) and focus filters (web, academic, social, finance, video). Migrate ChatGPT 'Search' users to Perplexity's default mode; migrate users who run citation-heavy research workflows to Pro Search or Deep Research.
- Warning: ChatGPT Search is one mode among many (chat, code, image, agent) and is not always on; Perplexity Search is the default for every thread, which changes the prompting style users expect.
- Warning: Perplexity exposes a much richer mode/focus matrix - users coming from ChatGPT need to learn the Pro/Deep Research/Labs distinction or they'll over-spend their daily caps.
- Warning: Both products cite sources but Perplexity's citation density and click-through UI is markedly more aggressive; ChatGPT may give one or two links per answer where Perplexity gives a numbered list.