Perplexity

An AI search engine that answers with source links

SearchChatbot/Assistant

CurrencyUSDKRWEUR

Personal starting price$20+Official price in USD
Free tierFree plan available
Korean supportGood

Price checked 2026-10-19, source: perplexity.ai/help-center/en/articles/11680686-perplexity-max. Prices and benefits may change. Check the official site before you pay.

Pros

  • Sources shown for every answer
  • Strong at searching the latest information
  • Can pick models from several companies

Cons

  • Weaker than general chatbots at long creative writing and conversation
  • Source summaries are sometimes inaccurate

These are opinions compiled by Pickx from public sources.

At a glance

Perplexity is an AI search engine that searches the web in real time when you ask a question and organizes the answer with source links. In addition to its own Sonar models, you can choose the latest models from several companies, such as GPT, Claude, Gemini and Grok, so it suits office workers and students who often research the latest information and organize materials. Recently it has also added the AI browser Comet, Computer, which carries out tasks for you, and an in-depth investigation (Research) feature.

Full model list

Below is the search model list from the Perplexity Help Center (as of 2026-09-04). The actual list for each account follows the app's model picker, and models are added and retired from time to time.

ModelTypeFeaturesAvailable plans
Best (automatic selection)Automatic routing
  • Perplexity automatically picks the model that fits the question
  • Free users can only use this mode
All plans including Free
Sonar 2Perplexity's own model
  • In-house model specialized for search answers
  • Fast
  • Source citations are fairly accurate
  • Pro
  • Max
  • Enterprise Pro/Max
GPT-5.6 TerraOpenAI
  • OpenAI's main general-purpose model
  • Strong across writing, analysis and reasoning
  • Pro
  • Max
  • Enterprise Pro/Max
GPT-5.6 SolOpenAI (higher-tier model)
  • OpenAI's top reasoning model
  • Suited to hard analysis and expert questions
  • Max
  • Enterprise Pro/Max
Gemini 3.7 FlashGoogle
  • Google model
  • Fast responses
  • Handles long context
  • Pro
  • Max
  • Enterprise Pro/Max
Claude Sonnet 5Anthropic
  • Anthropic's main model
  • Strong at natural writing
  • Strong at code and document work
  • Pro
  • Max
  • Enterprise Pro/Max
Claude Opus 5Anthropic (higher-tier model)
  • Anthropic's top model
  • Suited to complex reasoning and long document analysis
  • Max
  • Enterprise Pro/Max
Kimi K3Moonshot AI
  • Model from China's Moonshot AI
  • Strong at long-context handling and reasoning
  • Pro
  • Max
  • Enterprise Pro/Max
GLM 5.3Z.aiOpen-family model from Z.ai (Zhipu) / Good-value general answers
  • Pro
  • Max
  • Enterprise Pro/Max
Grok 4.6xAI
  • xAI's model
  • Strong on real-time topics
  • Blunt answer style
  • Pro
  • Max
  • Enterprise Pro/Max
Nemotron 3 UltraNVIDIA
  • NVIDIA's open model
  • Large model tuned for reasoning and tool use
  • Pro
  • Max
  • Enterprise Pro/Max
Model CouncilCompare several models at once
  • Asks the same question to several top models and compares/combines the answers
  • Web only
  • Max
  • Enterprise Max
  • Comet (browser assistant) and Computer (task agent) modes use a separate model picker, so they may differ from the list above.
  • On Enterprise, admins can block certain companies or models, so the list may be shorter.
  • Image and video generation models (for example Sora 2 Pro video on Max) are offered by plan, but the exact model lineup changes often and we could not confirm it.

Summary by plan

PlanPrice (monthly billing / annual billing)What you can doUsage limitsFeatures
Free (Standard)Free
  • Basic search
  • Pro search
  • Research
  • File upload
  • Cannot choose models directly or generate images
  • Basic search effectively unlimited
  • Pro search 3 per day
  • Research 1 per month
  • File upload limited
Good for trying out searches with sources
Pro
  • $20/mo
  • $200/yr (about $17/mo)
  • Pro search/Research/Comet assistant/file and app creation
  • Advanced model selection
  • Image and video generation
  • Extended use of Computer
  • 'Normal use' level weekly/monthly limits for Pro search, Research, Comet, and file and app creation
  • 50 files per project
  • Can opt out of your data being used for AI training
  • Priority support
  • Korean won price cannot be confirmed officially (may differ when paying through app stores)
Education Pro
  • $10/mo
  • Requires SheerID student/faculty verification
  • All Pro features
  • Learn Mode
  • Computer
Unlimited Pro searchDiscounted plan for verified students and educators only
Max
  • $200/mo
  • $2,000/yr (about $167/mo)
  • Top models such as GPT-5.6 Sol/Claude Opus 5
  • Model Council
  • Computer
  • Sora 2 Pro video
  • Highest limits on all features
  • 10,000 Computer credits per month
Early access to new features (such as the Brain research preview) / Dedicated support / Annual billing available only on the web (perplexity.ai)
Enterprise Pro
  • $40/mo per seat
  • With annual billing $34/mo per seat ($400/yr)
  • Pro search
  • Research
  • Comet
  • File and app creation
  • File upload
  • Seat management
  • Shared projects
  • File connectors
  • Pro search 400 per week
  • Research 50 per month
  • Comet queries 80 per month
  • File and app creation 50 per month
  • File upload 100 per week
  • Not used for data training
  • Educational institutions and nonprofits $30/mo per seat ($300/yr)
Enterprise Max
  • With annual billing about $271/mo per seat
  • Monthly billing price not shown on the official page
  • Enterprise Pro features
  • Video generation
  • My Files
  • SCIM, audit logs, data retention settings, Insights dashboard and more
  • About 10x the limits of Enterprise Pro
  • Pro search 4,000 per week
  • Research 500 per month and more
  • 15 videos of 8 seconds per month
  • 10,000 My Files
  • Security and governance features
  • Some security features require a separate contract
API PlatformUsage-based pay-as-you-go
  • Sonar API: sonar input/output $1 per 1,000,000 tokens
  • sonar-pro input $3/output $15 + $5 to $14 per 1,000 requests
  • Search API $5 per 1,000 requests (Fast $1)
  • Agent API can also call outside models
  • No free credits
  • Pro subscribers get $5 in monthly API credits
  • For developers
  • Does not include the web/app Pro features

Which plan for whom

  • Free: People who occasionally want to find the latest news or fact-check quickly with sources. Basic search is generous enough to replace everyday searching.
  • Pro: People who research materials every day for work or study and want to switch between GPT, Claude and Gemini as needed. It is like using several companies' models on one subscription, so the value is good.
  • Education Pro: People who can verify as students or faculty. You get Pro features at half price, with unlimited Pro search.
  • Max: Professionals and researchers who run deep research all day, use top models such as GPT-5.6 Sol/Claude Opus 5, or automate lots of work with Computer.
  • Enterprise Pro: Companies that adopt it team-wide, need internal data kept out of training, and must manage seats and permissions.
  • Enterprise Max: Large companies and professional organizations with very heavy research use that need security and governance such as SCIM and audit logs.
  • API Platform: Developers who want to add a 'search answer with sources' feature to their own service or internal tools.

Notes

Good for

#AI#Artificial Intelligence#Generative AI#Search#Chatbot#Assistant#AI Assistant#Research#Free

5 views

Prices converted at 1 USD = ₩1,329.49 (2026-10-19). Original price shown in brackets. Taxes as listed by each provider.

Comments

No comments yet. Be the first.

Your comments and nickname are public. If you see abuse or advertising, report it here and we will review and remove it.