"A free AI-powered answer engine that provides accurate, trusted, and real-time answers to any question."
Perplexity is an AI-powered search and answer platform that combines real-time web search with summarized answers and source citations.
Today, the product portfolio includes not only Search/Pro Search/Deep Research, but also Internal Knowledge Search, Connectors, asset creation for documents/presentations/spreadsheets/apps, browser/agent capabilities, as well as its own API platform.
Perplexity
A free AI-powered answer engine that provides accurate, trusted, and real-time answers to any question.
Location: USA ⓘ San Francisco, California
Perplexity Max: Includes Pro features plus the highest access to advanced AI models, expanded limits for “Create files and apps,” early access to new products, and prioritized support.
Education Pro: For verified students and educators; includes Pro features plus more citations, file and image uploads, Study Mode, expanded access to Perplexity Research and Perplexity Academic.
Enterprise Pro: Includes Pro features plus strict privacy features, no use of data for training, seat/user management, organizational file storage, internal knowledge search, increased file uploads, admin controls, team spaces, and dedicated support.
Enterprise Max: Includes Enterprise Pro features plus higher limits for Research, file and app creation, video generation, advanced models, larger file repositories, SCIM, audit logs, configurable data retention, and an insights dashboard. Other Sonar API: Separate pay-as-you-go model for developers and businesses that want to integrate Perplexity features into their own products, workflows, or automations. API usage is managed separately via credits or auto top-up; a Perplexity subscription is not required for API access. The API does not offer Pro/Max features within the web app and, according to Perplexity, does not log or store data.
Important: API access is not automatically included in Enterprise plans; Enterprise and API are treated separately.
Target audience
Perplexity is aimed at several clearly distinguishable target groups: first, private users and knowledge workers who want a faster, citable alternative to traditional web search; second, freelancers, analysts, consultants, and editorial teams who create more complex research, comparisons, and reports; third, organizations that want to combine internal knowledge from files and connected tools with web sources; fourth, developers who want to integrate Perplexity search or Sonar models into their own products. Education Pro and Learn Mode also show that education and academic use have now become an official focus.
Outstanding features
The strongest differentiating features are the combination of up-to-date web search, answers with sources, Deep Research, Internal Knowledge Search, and asset creation. Unlike pure chat tools, Perplexity combines research with operational follow-up processing: documents, presentations, spreadsheets, or small apps can be generated from a research thread. For teams, there are also org repositories, connectors, admin controls, retention settings, and auditing; for developers, Search API, Agent API, and Sonar API.
Main areas of application
Perplexity is strongest wherever current, verifiable information needs to be quickly turned into usable results: market and competitive research, academic research, due diligence, internal knowledge search, report creation, presentation preparation, source collection, file/document analysis, and research-oriented knowledge work. In the enterprise sector, the focus also shifts further toward an internal knowledge hub, team collaboration, connector-based search, and governance. Via the API, Perplexity is also suitable for search, research, and agent functions in proprietary applications.
Usage & notes
For simple questions, Free is often sufficient; for reliable work use, Pro is the more realistic starting point. Anyone working with confidential data, internal documents, or team knowledge should not equate the consumer plans with Enterprise: with Free/Pro/Max, AI data usage is generally opt-out by default, whereas Enterprise explicitly offers “never used for training” and more controls. It is also important to check pricing carefully, because Perplexity now combines several product lines: Consumer Search, Research, Create, Comet/Computer, Enterprise, and API. This is precisely what makes the platform powerful, but also less lightweight than a pure search engine.
| Who is it suitable for? | Assessment & rationale |
|---|---|
| Private individuals | Suitable – Perplexity offers a Free/Standard plan with many basic searches and limited Pro searches; useful for quick everyday research, learning, topic overviews, and finding sources. For intensive use, Pro or Max are more suitable because they include advanced models, more file uploads, as well as image and video generation. |
| Self-employed / freelancers | Very suitable – especially for research, content ideas, market and competitive analysis, SEO preparation, client briefings, and quick summaries. This aligns strongly with the use cases Research, Texts / Content, SEO / GEO, Marketing / Advertising, and Knowledge management / internal search. |
| SMEs / small teams | Suitable to very suitable – Perplexity Enterprise is designed for teams that want to use web research, internal files, and tools on one platform; according to Perplexity, teams can retrieve answers from files, integrations, and web sources. |
| Large enterprises | Very suitable – according to Perplexity, Enterprise Pro is aimed at organizations that want to increase productivity, answer complex questions faster, and meet stricter security requirements; it offers, among other things, user management, higher data protection standards, and SOC 2 alignment. |
| Developers / technical teams | Very suitable – the Perplexity API can be used for Search, Sonar, Agent API, and embeddings; suitable for AI search, RAG, Q&A applications, research tools, AI assistants, and apps with current web sources. |
| Privacy-conscious companies | Conditionally to well suited – positive: Perplexity states that it does not use Enterprise data for training or fine-tuning; according to the Help Center, uploaded files are deleted after seven days. Important: with Free, Pro, and Max, “AI Data Retention” is enabled by default and must be actively disabled if data is not to be used for training. |
| Pure creative/image users | Conditionally suitable – according to Perplexity, Pro offers image and video generation, but Perplexity’s core strength clearly lies in search, research, source work, and knowledge questions, not primarily in creative production. |
| Users who only need chat without sources | Conditionally suitable – Perplexity can be used in a chat-like way, but it is particularly strong when current information, sources, web research, and verifiable answers matter. For pure writing assistance without research, ChatGPT, Claude, or Gemini may be more suitable depending on the workflow. |
Hosting & Data
1) On-prem / local hosting
Meaning: The company operates the solution on its own hardware or within its own infrastructure. In the strictest sense, not only the application runs locally, but ideally the model as well.
2) Private cloud / data center
Meaning: The solution runs in a dedicated or more clearly separated cloud environment, often with a hosting provider or hyperscaler, but in a German data center or in a particularly controlled environment.
3) EU SaaS / managed
Meaning: The provider operates the solution itself as a service. The company uses the tool as a ready-made cloud service, ideally with EU data residency.
4) Hybrid
Meaning: One part of the processing remains internal / local / in a private cloud, while another part runs in an external cloud or EU SaaS.
5) AVV / DPA
Meaning: This is the data processing agreement or Data Processing Addendum. It governs that the provider processes personal data on behalf of the customer and is bound by the customer's instructions.
6) No training
Meaning: The provider does not use your prompts, uploads, attachments, chat histories, or outputs for training or improving the general model — ideally excluded by contract.
7) Open-source / transparency path
Meaning: There is a path toward greater technical transparency and sovereignty, for example through:
- open models
- documented components
- self-hostable parts
- traceable architecture
- export / switching options
| On-prem / local hosting | ❓ |
| Private cloud / data center | ⚠️ |
| EU SaaS / Managed | ⚠️ |
| Hybrid | ❓ |
| DPA / AVV | ✅ |
| No training on customer data | ⚠️ |
| Open source / transparency path | ⚠️ |
On-Prem / local hosting: indirect / not available
On-prem, local hosting, or self-hosting is not specified on the website. The documented offerings refer to SaaS, Enterprise, and API.
Private Cloud / data center: partial
There are Enterprise and API offerings with additional data protection controls and for the Sonar API Zero Data Retention. However, a dedicated EU/EEA private cloud or a clearly designated segregated EU data center is not specified on the website.
EU SaaS / Managed: partial
Perplexity is generally addressable in the EU/EEA region, names an EU representative, and mentions transfer mechanisms. However, explicit EU data residency or an EU/EEA operating region for the managed SaaS service is not specified on the website; the privacy policy primarily refers to the USA.
Hybrid: unclear
A hybrid operating model with an internal/local component plus an external service is not specified on the website.
DPA / DPA: covered
A DPA is published on the website. In it, Perplexity is described in the relevant context as a processor, with provisions on subprocessors, support, security measures, audits, as well as EU/UK data protection references.
No training: partial
For Enterprise, it is stated that data is never used for AI training. For the Sonar API, according to the Privacy & Security page, Zero Data Retention applies and customer data is not used for training. However, for Free, Pro, and Max, training is enabled by default and must be manually disabled via opt-out; according to the website, previously collected training data cannot be deleted or removed.
Open-source / transparency path: partial
A certain transparency path is documented: Perplexity lists third-party models and states that Sonar is based on the open-source model Llama 3.1 70B and is fine-tuned by Perplexity. However, a comprehensive open-source, self-hosting, or openly documented component path for the overall product is not specified on the website.
Data processing
The website describes several processing paths: general service use with possible storage/processing mainly in the USA; Enterprise with no-training commitments and additional retention/admin controls; Sonar API with Zero Data Retention and without use of customer data for training. The DPA names AWS as the global cloud infrastructure and, regarding model providers, refers to a separate third-party model list. Fixed EU/EEA data residency for the entire service is not specified on the website.
Conclusion
For users in the European region, Perplexity is not documented as a clearly fully EU/EEA-resident standard SaaS. More GDPR-aligned use is under certain conditions more readily demonstrable via Enterprise or Sonar API, supported by DPA, SCC/DPF references, no-training or zero-data-retention statements, and EU representation. Due to primary processing in the USA, lack of demonstrated EU data residency, and training data use being enabled by default for consumer plans, the overall assessment for the EU/EEA region is 'conditional'.
Sources
- https://www.perplexity.ai/help-center/de/articles/11564572-datenerfassung-bei-perplexity
- https://www.perplexity.ai/es/hub/legal/privacy-policy
- https://www.perplexity.ai/es/hub/legal/dpa
- https://www.perplexity.ai/ja/hub/legal/third-party-models
- https://docs.perplexity.ai/docs/resources/privacy-security
| On-prem / local hosting | ❓ |
| Private cloud / data center | ⚠️ |
| EU SaaS / Managed | ⚠️ |
| Hybrid | ❓ |
| DPA / AVV | ✅ |
| No training on customer data | ⚠️ |
| Open source / transparency path | ⚠️ |
On-Prem / local hosting: indirect / not available
On-prem, local hosting, or self-hosting is not specified on the website. The documented offerings refer to SaaS, Enterprise, and API.
Private Cloud / data center: partial
There are Enterprise and API offerings with additional data protection controls and for the Sonar API Zero Data Retention. However, a dedicated EU/EEA private cloud or a clearly designated segregated EU data center is not specified on the website.
EU SaaS / Managed: partial
Perplexity is generally addressable in the EU/EEA region, names an EU representative, and mentions transfer mechanisms. However, explicit EU data residency or an EU/EEA operating region for the managed SaaS service is not specified on the website; the privacy policy primarily refers to the USA.
Hybrid: unclear
A hybrid operating model with an internal/local component plus an external service is not specified on the website.
DPA / DPA: covered
A DPA is published on the website. In it, Perplexity is described in the relevant context as a processor, with provisions on subprocessors, support, security measures, audits, as well as EU/UK data protection references.
No training: partial
For Enterprise, it is stated that data is never used for AI training. For the Sonar API, according to the Privacy & Security page, Zero Data Retention applies and customer data is not used for training. However, for Free, Pro, and Max, training is enabled by default and must be manually disabled via opt-out; according to the website, previously collected training data cannot be deleted or removed.
Open-source / transparency path: partial
A certain transparency path is documented: Perplexity lists third-party models and states that Sonar is based on the open-source model Llama 3.1 70B and is fine-tuned by Perplexity. However, a comprehensive open-source, self-hosting, or openly documented component path for the overall product is not specified on the website.
Data processing
The website describes several processing paths: general service use with possible storage/processing mainly in the USA; Enterprise with no-training commitments and additional retention/admin controls; Sonar API with Zero Data Retention and without use of customer data for training. The DPA names AWS as the global cloud infrastructure and, regarding model providers, refers to a separate third-party model list. Fixed EU/EEA data residency for the entire service is not specified on the website.
Conclusion
For users in the European region, Perplexity is not documented as a clearly fully EU/EEA-resident standard SaaS. More GDPR-aligned use is under certain conditions more readily demonstrable via Enterprise or Sonar API, supported by DPA, SCC/DPF references, no-training or zero-data-retention statements, and EU representation. Due to primary processing in the USA, lack of demonstrated EU data residency, and training data use being enabled by default for consumer plans, the overall assessment for the EU/EEA region is 'conditional'.
Sources
- https://www.perplexity.ai/help-center/de/articles/11564572-datenerfassung-bei-perplexity
- https://www.perplexity.ai/es/hub/legal/privacy-policy
- https://www.perplexity.ai/es/hub/legal/dpa
- https://www.perplexity.ai/ja/hub/legal/third-party-models
- https://docs.perplexity.ai/docs/resources/privacy-security
Strengths & weaknesses at a glance
| Strengths | Weaknesses |
|---|---|
| • Very strong fit for research with source citations and up-to-date web information. | • For Free/Pro/Max, AI data usage or data usage for model improvement is enabled by default until the user actively opts out; according to the Help Center, earlier training data is not removed retroactively from previously collected training data. |
| • Deep Research for complex, multi-step analyses. | • Truly strong privacy/governance features are only available in Enterprise. |
| • Internal Knowledge Search combines web and internal files/repos. | • The public pricing is partly inconsistent: the Enterprise landing page shows annual prices converted to 34/271 USD per seat and month, while the Help Center lists 40/325 USD monthly or 400/3,250 USD annually. |
| • Create files and apps for documents, PPTX, XLSX, and HTML directly from prompts. | • Max/Enterprise Max are positioned significantly higher in terms of pricing. |
| • Broad connector/API ecosystem for developers and teams. | |
| • Enterprise with SSO/SCIM, audit logs, data retention, admin controls. |
Reviews
1 review in total
Submit review
Your review will only become visible after email confirmation. This protects the portal against abuse.
Report review
Please select the reason why this review should be checked.
GDPR-compliant usage possible?
Perplexity provides several components relevant for GDPR-compliant use on its own website, including a privacy policy, a DPA, EU representation, SCC/DPF notices, as well as opt-out and no-training statements for certain products. However, for use in the EU/EEA, the situation is not entirely straightforward: the privacy policy identifies the USA as the primary storage and processing location, explicit EU data residency for general SaaS use is not stated on the website, and for Free/Pro/Max, training use is enabled by default until users opt out. A relatively privacy-friendly path is documented for Enterprise as well as for the Sonar API with Zero Data Retention, but this depends on the product or plan and is not equivalent to generally documented EU/EEA data residency.
Positive
Positive aspects include the published privacy policy, an accessible DPA, identified subprocessor/third-party model structures, EU representation in Berlin, notices regarding SCCs and DPF for EU/UK transfers, a documented opt-out for AI training for Free/Pro/Max, the statement that Enterprise data is never used for AI training, and, for the Sonar API, a strict Zero Data Retention policy. In addition, the website mentions a SOC 2 Type II certification.
Negative
Negative aspects are that the privacy policy describes data processing primarily in the USA and does not state explicit EU/EEA data residency for the general service. For Free, Pro, and Max, AI data storage or data collection for training is enabled by default; an opt-out applies only to future data, and according to the Help Center, previously collected training data cannot be deleted or removed. Server/data center locations in the EU or dedicated EU regions are not stated on the website. On-prem or self-hosting options are not stated on the website.
Server location
According to the privacy policy, data may be stored or processed in other countries, 'primarily in the USA, where we have our servers and support teams.' A specific EU/EEA server location or EU data residency is not stated on the website. The DPA identifies Amazon Web Services as the cloud infrastructure provider with location 'worldwide.'
