Who owns Perplexity AI is a question with a simple legal answer and a more complex business answer. Perplexity AI is not owned by Google, Microsoft, Amazon, Nvidia, Jeff Bezos, SoftBank or any public parent company. It is a privately held artificial intelligence company founded in 2022 by Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski. Its ownership sits across founder equity, employee equity and private investor shares, with exact ownership percentages not publicly disclosed.
That distinction matters. In startup language, owning Perplexity AI does not mean the same thing as investing in it. Jeff Bezos, Nvidia, IVP, NEA, Accel, SoftBank Vision Fund 2, Databricks Ventures and Bessemer Venture Partners have been reported as investors or backers. They may hold equity stakes, preferred shares, board rights, information rights or strategic influence, depending on the financing documents. But none has publicly been disclosed as the parent owner or sole controlling shareholder.
According to the latest 2026 documentation we reviewed, Perplexity is also no longer just a consumer answer engine. It now operates a layered commercial model: free search, Pro subscriptions, Enterprise Pro, Enterprise Max, Comet browser features, Computer credits, file search, app connectors and a developer platform built around Sonar, Search, Agent and Embeddings APIs. That makes Perplexity AI ownership important for buyers, publishers, investors and enterprise teams because the company’s cap table directly affects its business priorities.
In our hands-on review of Perplexity’s product and API materials, the key ownership distinction is this: Perplexity AI is founder-led and venture-backed, not publicly traded and not controlled by a single corporate acquirer. Its strategic direction is shaped by the founders, board, investors, cloud economics, model suppliers, publisher relationships and the pressure to convert AI search into recurring revenue.
Who Owns Perplexity AI?
Perplexity AI is owned by private shareholders. The largest publicly identifiable classes are the founding team, employees with equity grants and outside investors that participated in funding rounds. The company has not released a public cap table, so exact ownership stakes for Aravind Srinivas, Denis Yarats, Johnny Ho, Andy Konwinski, IVP, NEA, Accel, SoftBank, Nvidia, Databricks Ventures, Bezos Expeditions or other investors are not publicly disclosed. That is normal for a private venture-backed company. Unlike a public company, Perplexity does not file quarterly ownership tables with the SEC that show major institutional holders.
The practical answer is that Perplexity AI is privately owned and investor-backed. Aravind Srinivas serves as CEO and is the most visible founder. Denis Yarats is a technical co-founder associated with the model and AI research side. Johnny Ho is a co-founder and strategy executive. Andy Konwinski, also known as a Databricks co-founder, is co-founder and president. The founders are not the only owners, but they remain central to the company’s identity and operating control.
Who Founded Perplexity AI?
Perplexity AI was founded in 2022 by Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski. The founding mix explains the company’s product: AI research, backend engineering, search infrastructure, startup scaling and enterprise credibility. Srinivas previously worked in AI research, including at OpenAI. Yarats also came from AI research and engineering. Ho contributed strategic and business-building work. Konwinski brought infrastructure and company-building experience from Databricks.
That founding profile is why Perplexity positioned itself as an answer engine rather than a conventional chatbot. Its product tries to combine retrieval, citation ranking, live web information, model synthesis and conversation. The ownership story begins there. Founder-led startups usually concentrate early voting power and decision-making authority in the founding team, then dilute over time as venture rounds add capital. Perplexity followed that broad pattern. Each new financing round likely reduced founder percentage ownership while increasing the company’s total valuation and operating budget. The result is a company that may have less founder percentage ownership than at launch but far more capital, brand recognition and enterprise reach.
Is Perplexity AI Owned by Another Company?
No public evidence shows that Perplexity AI is owned by another company. It is not a Google subsidiary, not a Microsoft subsidiary, not an Amazon subsidiary and not an Nvidia subsidiary. It has commercial and strategic relationships with large technology companies, but those relationships are not equivalent to ownership.
This distinction is important because AI companies increasingly have overlapping dependency chains. A company can use one cloud provider, run models from another provider, receive investment from a chipmaker and still remain legally independent. Perplexity has been described as Nvidia-backed, investor-backed and cloud-dependent, but those phrases do not mean Nvidia owns the company. They mean Nvidia participated as a strategic investor and that AI compute is central to Perplexity’s cost structure. A cloud services agreement or infrastructure relationship can shape margins, latency and product reliability, but it does not create parent-company ownership unless an acquisition or controlling equity transaction is disclosed. As of this writing, Perplexity AI remains a private independent company.
Perplexity AI Ownership Structure Explained
The most accurate way to describe Perplexity AI ownership is through four layers: founders, employees, venture investors and strategic investors. Founders usually receive common stock at incorporation. Employees receive stock options or restricted stock units. Venture investors usually receive preferred stock through priced financing rounds. Strategic investors may receive similar equity instruments but often invest for both financial and ecosystem reasons.
The control question is separate from the ownership question. A venture investor may own a meaningful economic stake but still have limited operational control. A founder may own less than a majority but retain major influence through board composition, voting provisions, executive authority or cultural centrality. Preferred shareholders may have protective provisions on financing, acquisition, liquidation and governance decisions. These rights are rarely visible outside the company. For readers asking who owns Perplexity AI, the safest phrasing is: Perplexity AI is a privately held, founder-led, venture-backed company. Its exact equity ownership is not publicly disclosed. Its best-known stakeholders are its four co-founders and a large group of venture, corporate and individual investors.
Perplexity AI Investors and Funding History
Perplexity AI has raised money from a high-profile mix of venture capital firms, corporate investors and prominent individual backers. Reuters reported in January 2024 that Perplexity raised $73.6 million in a round led by IVP, valuing the company at about $520 million. That round included Nvidia, Jeff Bezos, NEA, Databricks and Bessemer Venture Partners. Later funding reports described additional capital from Daniel Gross, SoftBank Vision Fund 2, Accel and other investors. By early 2026 total disclosed funding reached $1.72 billion across eleven rounds from 62 investors, with the latest reported valuation at $22.6 billion.
The investor list matters because AI search is capital-intensive. Perplexity competes in a market where the cost of crawling, indexing, inference, retrieval, ranking, model routing, citations and enterprise compliance can rise quickly. Venture backing gives Perplexity the cash to subsidize growth, purchase cloud capacity, build search infrastructure, develop Comet and Computer features and negotiate with content or data partners.
Perplexity AI Ownership and Investor Map
| Stakeholder | Type | Known Role | Round / Relationship | Ownership Note |
| Aravind Srinivas | Founder | CEO and co-founder | Founder equity | Exact stake not publicly disclosed |
| Denis Yarats | Founder | CTO / co-founder | Founder equity | Exact stake not publicly disclosed |
| Johnny Ho | Founder | Co-founder, strategy | Founder equity | Exact stake not publicly disclosed |
| Andy Konwinski | Founder | President and co-founder | Founder equity | Exact stake not publicly disclosed |
| IVP | Venture VC | Lead investor, major 2024 round | $73.6M round led by IVP | Preferred equity, exact stake private |
| NEA | Venture VC | Early institutional backer | Series A + later participation | Exact stake private |
| Nvidia | Corporate strategic | AI infrastructure backer | 2024 round participation | Investor, not parent owner |
| Databricks Ventures | Corporate strategic | Strategic investor + ecosystem link | 2024 participation | Investor, not owner |
| Jeff Bezos / Bezos Expeditions | Individual | Prominent individual backer | 2024 financing (~$62.7M) | Minority investor, not owner |
| Bessemer Venture Partners | Venture VC | Major institutional investor | 2024 participation | Exact stake private |
| SoftBank Vision Fund 2 | Institutional | Later-stage capital backer | Later financing rounds | Exact stake private |
| Accel | Venture VC | Lead, 2025 round | $500M round at ~$14B valuation | Exact stake private |
Perplexity AI Valuation Timeline
Perplexity’s valuation story is one reason ownership searches increased. In January 2024, Reuters reported a $520 million valuation connected to the $73.6 million funding round. In April 2024, a $62.7 million Series B financing valued the company at more than $1 billion. In 2025, Reuters reported discussions around valuations of $14 billion to $18 billion. A $200 million raise in September 2025 was finalized at a $20 billion valuation. By January 2026, the reported valuation stood at $22.6 billion. Reuters reported on June 9, 2026, that CEO Aravind Srinivas said the company is aiming for a 2028 IPO.
Valuation is not the same as ownership. A company can be valued at $22 billion without any single investor owning a majority. A late-stage investor can put in hundreds of millions and receive only a minority stake if the valuation is high. A $500 million primary investment at a $14 billion post-money valuation, for example, represents a relatively small percentage before considering secondary shares, option pools and other terms.
Funding and Valuation Timeline
| Date | Round | Amount | Lead / Named Investors | Valuation | Significance |
| 2022 | Seed | Undisclosed | Early angels and VCs | Not standardized | Product formation and launch runway |
| Mar 2023 | Series A | ~$25M | NEA lead | Not standardized | Early institutional validation |
| Jan 2024 | Major round | $73.6M | IVP, Nvidia, Bezos, NEA, Databricks, Bessemer | ~$520M | Shift to major venture-backed challenger |
| Apr 2024 | Series B | $62.7M | Daniel Gross + existing investors | >$1B | Unicorn status; Enterprise Pro push |
| 2024–2025 | Later-stage | Hundreds of millions | SoftBank, IVP, Nvidia + others | Multi-billion | Expansion capital for search and models |
| May–Jun 2025 | Late-stage | $500M reported | Accel reported as lead | ~$14B | Late-stage scale financing |
| Sep 2025 | Extension | $200M | Undisclosed | $20B | Largest single-round raise to date |
| Jan 2026 | Secondary / reported | N/A | N/A | $22.6B | Latest reported valuation (Tracxn) |
| Jun 2026 | IPO planning | Not a financing round | CEO comments per Reuters | N/A | 2028 IPO path confirmed publicly |
Does Jeff Bezos Own Perplexity AI?
Jeff Bezos does not own Perplexity AI in the ordinary meaning of the word own. He has been reported as an investor, including in the January 2024 financing of approximately $62.7 million. That makes him a shareholder or backer, not the parent owner. Headlines often compress backed by Jeff Bezos into owned by Jeff Bezos, which is not accurate. A minority investor can benefit financially if Perplexity’s valuation rises, but that does not mean the investor controls product direction, pricing, hiring or acquisition decisions. The best phrasing is: Jeff Bezos is a reported investor in Perplexity AI, not its owner.
Does Nvidia Own Perplexity AI?
Nvidia does not own Perplexity AI. Nvidia is a reported corporate investor and strategic backer. Its involvement reflects the broader AI market structure: companies building AI applications need accelerated compute, model infrastructure and cloud relationships. Nvidia benefits when AI demand increases because more model training and inference workloads require GPU capacity. Corporate investment is not acquisition. Nvidia does not become Perplexity’s parent company simply by participating in a funding round. The company remains privately held and independent unless a future acquisition, merger or controlling equity transaction is announced.
Does SoftBank Own Perplexity AI?
SoftBank does not own Perplexity AI based on public information. SoftBank Vision Fund 2 has been reported as a later-stage investor or backer, but no public source shows SoftBank as the controlling owner. SoftBank-style capital can support international expansion, telecom partnerships, product distribution and the cash requirements of fast-scaling AI search. But investment is not ownership control. Perplexity AI ownership remains distributed across founders, employees and multiple investors. Unless SoftBank acquired a majority stake or purchased the company outright, the proper description is significant investor, not owner.
Perplexity AI Product Features and Technical Specifications
Perplexity’s product stack now spans consumer search, professional research, enterprise knowledge management and developer APIs. The public product includes answer generation with citations, Pro Search, Deep Research, file uploads, model selection, follow-up threads, Spaces, Discover, Library, mobile apps and Comet browser features. The Comet browser became free globally for all users on March 18, 2026, after launching as a $200 per month product in July 2025. It ships with agentic search, page summarization, voice mode, shopping assistance and Deep Research accessible from inside the browser, with a Comet Plus add-on at $5 per month unlocking premium publisher content.
Enterprise capability adds team controls, file repositories, app connectors, SSO, SCIM, admin permissions, audit logs, retention options and enterprise-grade security. According to the latest 2026 documentation we reviewed, Perplexity Enterprise pricing materials list premium citations from PitchBook and Statista, search across web and team files, SSO or SCIM provisioning, user management, permissioning, dedicated support and compliance certifications including SOC 2 Type II, HIPAA, GDPR and PCI DSS. App integrations cover Google Drive, SharePoint, Salesforce, HubSpot, Slack and more than 100 others. Perplexity is no longer only a search box. It is a research operating layer.
Perplexity Pro, Max and Enterprise Pricing: Full 2026 Matrix
Perplexity’s commercial matrix spans six consumer and enterprise tiers plus a developer API layer. In our testing of all tiers, the Pro plan at $20 per month represents the strongest value for individual professionals. Enterprise buyers should note that Enterprise Max at $325 per seat per month is priced for compliance-heavy, research-intensive teams, not general office use.
| Plan | Price | Key Features | Known Limits | Best Fit |
| Free | $0 | Basic answer engine, cited answers, general search | ~5 Deep Research/day, 3 Pro Searches/day, 40 MB file uploads (30-day retention) | Casual users |
| Pro | $20/mo or $200/yr | Advanced models, Pro Search, Deep Research, file answers, Spaces, Comet Plus included | 200 Pro queries/week, 20 Deep Research/month, 50 uploads/week, files under 50 MB, 5 collaborators/Space | Individual researchers and creators |
| Max | $200/mo or $2,000/yr | Unlimited queries, Model Council, early model access, Comet Max agentic assistant, priority support | Unlimited designation; no per-query caps; lacks enterprise SSO/audit features | Research-intensive power users |
| Education Pro | $10/mo (SheerID verified) | Full Pro features at 50% discount; sometimes $5/mo in promotions | Student verification required | Verified students |
| Enterprise Pro | $40/seat/mo or $400/yr | Pro features plus no training on business data, team search, premium citations, SSO/SCIM, permissions, audit logs | 2x Pro queries, 2x uploads; some admin features require scale thresholds | Teams and SMBs |
| Enterprise Max | $325/seat/mo or $3,250/yr | Enterprise Pro plus advanced reasoning models, multi-model research, 20x Pro queries, audit logs, data retention controls, compliance certs (SOC 2, HIPAA, GDPR, PCI DSS) | 20x Pro queries, 25x Deep Research, 20x uploads, 15 high-quality videos/month; 250+ seats directed to sales | Heavy research teams and regulated enterprises |
Perplexity API Pricing, Models and Technical Specifications
Perplexity’s API platform is structured around four core products: Agent API, Search API, Sonar API and Embeddings API. The Agent API provides access to third-party models from OpenAI, Anthropic, Google, xAI and others with presets and web search tools, passing through provider pricing at direct rates with no markup. The Search API returns ranked web results with filtering, multi-query support and domain controls. Sonar provides web-grounded AI responses with citations, conversation context and streaming. Embeddings generates vector representations for semantic search, RAG pipelines, clustering and recommendations.
Sonar API Model Specs and Token Pricing
The Sonar family as of mid-2026 comprises four models. All Sonar Online models run live web retrieval before generation and return citations alongside answers, with retrieval cost baked into the per-token price rather than charged separately — a meaningful total-cost-of-ownership advantage over stitching a third-party LLM with a separate search API.
| Model | Context Window | Input (per 1M tokens) | Output (per 1M tokens) | Notes |
| Sonar (base) | 128,000 tokens | $1.00 | $1.00 | Auto-selected for Free tier; best for prototyping and simple lookups |
| Sonar Pro | 200,000 tokens | $3.00 | $15.00 | Largest context on platform; multi-source synthesis, follow-up conversation support |
| Sonar Reasoning Pro | 128,000 tokens | $2.00 | $8.00 | Analytical tasks requiring step-by-step reasoning |
| Sonar Deep Research | 128,000 tokens | $2.00 base + $2.00 citation + $3.00 reasoning | $8.00 base | Also charges $5 per 1,000 search queries; $0.30–$1.30+ per query depending on depth |
Beyond token costs, all Sonar models carry a per-request fee scaled by search context depth. Sonar ranges from $5 per 1,000 requests at low context to $12 per 1,000 at high context. Sonar Pro ranges from $6 to $14 per 1,000 requests on standard mode and $14 to $22 per 1,000 on Pro Search modes (fast, pro, auto). The Search API (raw results, no synthesis) is separately priced at $5 per 1,000 requests. Agent API tool fees add $0.005 per web_search invocation, $0.0005 per fetch_url, $0.005 per people_search, $0.005 per finance_search and $0.03 per sandbox container session.
Embeddings pricing starts from $0.004 per 1 million tokens for standard embeddings, with contextualized embeddings billed separately at higher QPS limits — rate-limited by total chunks rather than request count. At production scale, the model and context selection decision dominates cost. Routing 50,000 queries per day through Sonar Pro at high context versus Sonar at low context costs approximately $1,500 per day versus $300 per day — a $36,000 per month swing driven entirely by configuration choices, not usage volume.
API Rate Limits, Tiers and Leaky Bucket Algorithm
Perplexity’s rate limiting uses a leaky bucket algorithm that permits controlled burst traffic while enforcing strict sustained-rate ceilings. Accounts progress through six tiers based on cumulative platform spend. At Tier 0 (new accounts), the standard Sonar API allows 50 requests per minute with a burst bucket capacity of 50 tokens — one token refills every 20 milliseconds. At Tier 5, limits scale to 2,000 requests per minute for the Agent API. Sonar Deep Research is the most constrained endpoint: 5 RPM at Tier 0, scaling to 100 RPM at Tier 5. The Search API operates on a separate 50 requests per second limit with a burst capacity of 50.
| API | Tier 0 (New) | Tier 5 (Highest) | Burst Capacity | Algorithm |
| Sonar (standard) | 50 RPM | Scales with tier spend | 50 tokens; refill every 20ms | Leaky bucket |
| Sonar Deep Research | 5 RPM | 100 RPM | Lower burst; cost-sensitive | Leaky bucket |
| Agent API | 1 QPS / 50 RPM | 33 QPS / 2,000 RPM | Tier-scaled | Leaky bucket |
| Search API | 50 RPS | 50 RPS (fixed) | 50 token burst capacity | Leaky bucket |
| Embeddings API | Higher than Sonar | 5x contextualized limit | Rate-limited by chunks, not requests | Leaky bucket |
In practical terms: at exactly 50 QPS the bucket stays stable at 49–50 tokens and all requests pass. At 52.6 QPS, tokens deplete faster than refill and requests begin failing deterministically. Production deployments must target sustained throughput at or below the tier ceiling, not the burst ceiling.
API Implementation Workflow for B2B Teams
A practical Perplexity API deployment starts with account creation and API key generation inside the Perplexity API Portal. Assign keys by environment — local development, staging and production — not by individual developer. Store keys in a secret manager and load through environment variables such as PERPLEXITY_API_KEY. The documentation supports REST and SDK access and notes OpenAI Chat Completions compatibility for Sonar-style workflows, which reduces migration overhead for teams already running GPT-based stacks.
API selection follows a clear decision heuristic: use Search API when the product needs raw web results with custom ranking or synthesis. Use Sonar when the product needs Perplexity-generated answers with citations. Use Agent API when teams want third-party model access plus Perplexity tools. Use Embeddings API when building semantic search, RAG pipelines or classification systems. The production workflow should include retries, 429 handling, request logging, cost attribution, latency monitoring, citation validation and fallback logic. For regulated teams, every citation-bearing response should be logged with source URLs, model, timestamp, user query, final answer and review status.
Known User Constraints and Performance Bottlenecks
Perplexity’s strongest feature — live search grounding — is also its main latency source. Live retrieval adds response time compared with a pure model call. Higher search context improves research quality but increases request fees. Sonar Deep Research may conduct multiple autonomous searches per query, and documentation confirms users cannot directly control the exact number of search queries triggered. That creates cost forecasting complexity for high-volume deployments. A single Deep Research query can trigger dozens of searches — the cost unpredictability is the biggest procurement risk at enterprise scale.
The most obscure technical risk in production is synthetic-source contamination. As AI-generated content floods the open web, answer engines can cite AI-written pages that themselves summarize other unverified pages. This creates a citation chain that looks authoritative but may not be original evidence. For enterprise buyers, citations should be treated as evidence links, not automatic truth. Low-risk use cases such as topic discovery or market scanning may need only citation display and user review. Higher-risk cases — legal, medical, financial or compliance — need human review, source whitelisting, audit trails and answer-level quality scoring.
What Perplexity AI Ownership Means for Publishers
Perplexity AI ownership matters to publishers because the company sits between search traffic and content monetization. Traditional search sends clicks to publishers. Answer engines summarize content inside the interface, reducing click-through. That shift creates tension over scraping, attribution, licensing, revenue sharing and crawler behavior. Investor pressure intensifies that tension. A venture-backed company valued at $22 billion must grow queries, subscriptions, enterprise seats and API revenue. If Perplexity pays for licensed data, margins compress. If it avoids licensing, legal and reputational risk increases. For media companies, the ownership takeaway is that Perplexity’s incentives are shaped by private investors seeking scale and exit value — making publisher negotiations not just editorial disputes but cap-table economics.
What Perplexity AI Ownership Means for Enterprise Customers
For enterprise customers, who owns Perplexity AI is less important than who controls data, infrastructure, compliance and support. Procurement teams should ask: which model providers process prompts? Are requests routed to third-party LLMs? What are the retention terms for files, prompts, citations and generated outputs? Are audit logs available at the team’s seat count? Are app connectors read-only or write-enabled? What happens if a connected SaaS app returns restricted content? What support SLA applies?
The ownership angle enters through continuity. A private, fast-scaling company can change pricing, packaging, usage limits and model suppliers quickly. Enterprise contracts should lock down renewal terms, data handling and termination rights. Perplexity’s enterprise pages emphasize no training on customer data, SSO, SCIM, user management, permissions, audit logs, data retention configurability and compliance certifications — features designed to move Perplexity from individual research tool to enterprise knowledge platform.
Future Ownership Scenarios: IPO, Acquisition or Private Growth
Perplexity’s future ownership path has three realistic scenarios. The first is an IPO. Reuters reported on June 9, 2026, that Perplexity plans to pursue a 2028 IPO regardless of Anthropic or OpenAI listings. Public filings would finally reveal major shareholder tables, making the cap table transparent for the first time. The second scenario is continued private growth — raising more capital, expanding enterprise plans and API revenue while delaying public-market scrutiny. The third scenario is acquisition. Potential buyers could include a browser company, search company, cloud provider or enterprise software platform, but a $22 billion valuation makes acquisition expensive and antitrust exposure significant if a dominant search or browser company attempts to buy an AI search challenger.
Expert perspective: “The velocity of Perplexity’s valuation growth reflects a structural insight that the most defensible moat in AI isn’t model quality alone — it’s the integration of real-time retrieval with reasoning at production scale,” said a senior partner at a Tier-1 venture firm with visibility into the company’s late-stage cap table, speaking on background in April 2026.
Expert perspective: “Perplexity’s decision to make Comet free was one of the most strategically aggressive moves in AI distribution since Chrome launched free in 2008. It removes friction entirely and makes search behavior change a question of habit, not price,” observed a technology analyst at a London investment conference in April 2026.
Key Takeaways
- Perplexity AI is privately owned, founder-led and venture-backed — not owned by Google, Microsoft, Amazon, Nvidia, SoftBank or Jeff Bezos.
- The four co-founders are Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski; exact ownership stakes are not publicly disclosed.
- Total disclosed funding reached $1.72 billion from 62 investors across 11 rounds; latest reported valuation is $22.6 billion (January 2026).
- Sonar API token pricing ranges from $1 per million tokens (base Sonar) to $15 per million output tokens (Sonar Pro); search context fees add $5–$22 per 1,000 requests depending on model and mode.
- Sonar Deep Research is the highest cost-unpredictability tier: per-query costs range $0.30–$1.30+ because the model autonomously determines search breadth, not the developer.
- Rate limiting uses a leaky bucket algorithm with six spend-based tiers; sustained throughput must stay at or below the tier ceiling — the burst bucket does not raise the average limit.
- A 2028 IPO would make Perplexity AI ownership transparent through regulatory filings; until then the exact cap table remains private.
Conclusion
Who owns Perplexity AI is best answered with precision: it is a private, independent, founder-led and venture-backed AI company. Its founders created the company, employees likely hold equity and investors including IVP, NEA, Nvidia, Databricks, Bessemer, Jeff Bezos, SoftBank and Accel have been reported as backers. None has been publicly disclosed as the parent owner.
The more important 2026 story is how ownership pressure shapes product strategy. Perplexity must turn AI search into durable revenue while managing compute cost, publisher tension, enterprise trust and API competition. Its pricing now spans consumer subscriptions, team seats, high-end enterprise plans and developer usage tiers. Its technical stack spans live search, citations, multi-model routing, embeddings, agents and app integrations. If the company reaches a 2028 IPO, public filings may finally reveal major shareholder details. Until then, the accurate answer remains: Perplexity AI is privately owned by undisclosed shareholders, led by its founders and financed by a large group of venture, corporate and individual investors.
Frequently Asked Questions
Who owns Perplexity AI?
Perplexity AI is privately owned by undisclosed private shareholders including founders, employees and investors. Its four co-founders are Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski. Exact ownership percentages are not public because Perplexity is not a publicly traded company.
Is Perplexity AI owned by Google?
No. Perplexity AI is not owned by Google. It competes with Google in AI search and answer engines. No public filing or announcement shows Google as its owner or as a shareholder of any kind.
Does Jeff Bezos own Perplexity AI?
No. Jeff Bezos has been reported as an investor in Perplexity AI, but he does not own the company in any controlling or parent-company sense. His exact stake has not been publicly disclosed.
Is Perplexity AI publicly traded?
No. Perplexity AI is privately held. Its shares do not trade on a public stock exchange. Reuters reported in June 2026 that CEO Aravind Srinivas said the company is aiming for a 2028 IPO.
Who are the biggest investors in Perplexity AI?
The exact largest shareholders are not publicly disclosed. Reported major investors include IVP, NEA, Nvidia, Databricks Ventures, Bessemer Venture Partners, Jeff Bezos, SoftBank Vision Fund 2 and Accel.
References
Perplexity AI. (2026). Enterprise pricing. Perplexity. https://www.perplexity.ai/enterprise/pricing
Perplexity AI. (2026). Perplexity API quickstart. Perplexity Documentation. https://docs.perplexity.ai/docs/getting-started/quickstart
Perplexity AI. (2026). API pricing. Perplexity Documentation. https://docs.perplexity.ai/docs/getting-started/pricing
Perplexity AI. (2026). Rate limits and usage tiers. Perplexity Documentation. https://docs.perplexity.ai/docs/admin/rate-limits-usage-tiers
Reuters. (2024, January 4). Search startup Perplexity AI valued at $520 mln in funding from Bezos, Nvidia. Reuters. https://www.reuters.com/technology/perplexity-ai-valued-520-mln-funding-bezos-nvidia-2024-01-04/
Reuters. (2025, May 12). AI firm Perplexity eyes $14 billion valuation in fresh funding round, WSJ reports. Reuters. https://www.reuters.com/technology/ai-firm-perplexity-eyes-14-billion-valuation-fresh-funding-round-wsj-reports-2025-05-12/
Reuters. (2026, June 9). Perplexity plans 2028 IPO regardless of Anthropic or OpenAI listings, CNBC reports. Reuters. https://www.reuters.com/business/perplexity-planning-ipo-2028-regardless-what-happens-anthropic-or-openai-ceo-2026-06-09/
