The Perplexity AI investors list matters because Perplexity is no longer only a consumer AI search product. It is now a venture-backed answer engine with strategic investors across cloud, chips, venture capital, software infrastructure and internet distribution. The company’s cap table includes NEA, IVP, NVIDIA, Databricks, Bessemer Venture Partners, SoftBank Vision Fund 2, Accel and Jeff Bezos, with individual backers including Elad Gil, Nat Friedman, Bob Muglia, Paul Buchheit, Tobias Lütke and Dylan Field.
Perplexity’s funding story accelerated after its January 2024 Series B, when Reuters reported a $73.6 million round led by IVP at a valuation of about $520 million, with participation from NVIDIA, Jeff Bezos, NEA, Databricks and Bessemer Venture Partners. By late 2024, Reuters reported the company raising $500 million at a $9 billion valuation, again with IVP involved. In March 2025, Reuters reported talks to raise at an $18 billion valuation. As of early 2026, secondary-market estimates and Tracxn data place the figure between $21.2 billion and $22.6 billion, with total funding exceeding $1.5 billion across roughly eleven rounds.
According to the latest 2026 documentation we reviewed, Perplexity’s commercial thesis rests on three layers: a consumer answer engine, Pro and Max subscriptions, and a developer API platform built around Sonar, Sonar Pro, Sonar Deep Research, the Search API and the Agent API. The investor list is therefore not just a financing record. It is a map of how AI search is being positioned as a new interface for web retrieval, enterprise knowledge access, real-time citations and application-level reasoning.
Perplexity AI Investors List: Full Named Backers and Funding Context
Perplexity’s earliest disclosed institutional backers came from the traditional venture stack. NEA published its investment thesis in March 2023, describing Perplexity as an answer engine that could challenge traditional search behavior. That Series A period also included named investors such as Elad Gil, Nat Friedman, Bob Muglia and Paul Buchheit, according to contemporaneous funding coverage. The strategic importance of that group is clear: GitHub, Gmail, Microsoft, Databricks and AI infrastructure experience entered the company’s orbit before Perplexity became a high-valuation AI search company.
The Series B widened the cap table. Reuters reported that IVP led the $73.6 million round in January 2024, with NVIDIA, Jeff Bezos, NEA, Databricks and Bessemer Venture Partners participating. Fortune’s coverage also identified Shopify co-founder Tobias Lütke among investors in the same financing wave. This matters because Perplexity’s investor base moved from software-native venture backers into chip, cloud, commerce and founder-operator capital.
By 2025, the investor narrative became less about seed-stage validation and more about whether answer engines could claim part of Google’s search profit pool. Reuters reported a possible $18 billion valuation round in March 2025, noting support from NVIDIA, Jeff Bezos and SoftBank. In June 2026, Reuters reported that Perplexity is still targeting a 2028 IPO regardless of what happens with Anthropic or OpenAI listings.
Named investor matrix
| Investor / Backer | Type | Known Role | Strategic Signal |
| NEA | VC firm | Series A lead; later participant | Early answer-engine thesis |
| IVP | VC firm | Led Jan 2024 Series B; Nov 2024 round | Growth-stage validation |
| NVIDIA | Strategic corporate | Jan 2024 participant | AI compute & inference signal |
| Databricks | Strategic software | Early + Series B participant | Data infrastructure credibility |
| Bessemer Venture Partners | VC firm | Series B participant | Enterprise software network |
| Jeff Bezos / Bezos Expeditions | Individual / family office | Jan 2024 participant | Consumer internet & distribution |
| SoftBank Vision Fund 2 | Growth investor | Reported later-stage backer | Large-scale AI capital appetite |
| Accel | Growth VC | Reported 2025 financing coverage | Global venture scaling signal |
| Elad Gil | Individual angel | Early investor | AI & startup operator network |
| Nat Friedman | Individual angel | Early investor | Developer platform credibility |
| Bob Muglia | Individual angel | Series A period | Enterprise software credibility |
| Paul Buchheit | Individual angel | Series A period | Search, email & product design |
| Tobias Lütke | Individual angel | Reported 2024 backer | Commerce & product-led growth |
| Dylan Field | Individual angel | Reported 2024–25 backer | Design & developer tooling signal |
Sources: Reuters (Jan 2024, Nov 2024, Mar 2025), NEA blog (Mar 2023), Fortune (Jan 2024), Tracxn (2026). Later-round participants (Accel, SoftBank, Dylan Field) sourced from secondary-market and 2025 financing coverage.
Round-by-round funding and valuation ledger
The cleanest verified funding sequence is: NEA led the 2023 Series A, IVP led the January 2024 $73.6 million Series B at about $520 million, and IVP was also linked to the November 2024 $500 million round at a reported $9 billion valuation. Later figures around $14 billion, $18 billion and $20 billion are widely reported but less formally confirmed. For B2B diligence, treat post-$9B figures as reported valuations unless Perplexity directly confirms a round.
| Round / Date | Amount | Post-money Valuation | Lead / Key Investors |
| Pre-seed / Seed (2022–23) | ~$3.1M | ~$121M (Apr 2023) | NEA, Elad Gil, Nat Friedman, Databricks Ventures |
| Series A (2023) | ~$25.6M | ~$150M pre-money | NEA (Peter Sonsini), IVP, Bessemer, angels |
| Series B (Jan 2024) | ~$73.6M | ~$520M | IVP-led; NVIDIA, Bezos, Databricks, Bessemer, Lütke |
| Series C / Extension (Apr–Jun 2024) | $165M then $250M | $1B then $3B | SoftBank Vision Fund 2 (Jun 2024) |
| Series D (Nov–Dec 2024) | ~$500M | $9B | IVP-led, SoftBank |
| Extension (May 2025) | ~$500M | $14B | Existing investors |
| Extension (Jul 2025) | $100M | $18B | Undisclosed (Bloomberg) |
| Series D-IV (Sep 2025) | $200M | $20B | IVP + others (Reuters / The Information) |
| Angel tranche (Dec 2025) | Undisclosed | — | Incl. Cristiano Ronaldo (reported) |
| Series E-6 (early 2026) | Undisclosed | ~$21.2B–$22.6B | Secondary-market / Tracxn estimates |
Figures aggregate Reuters, Bloomberg, The Information, CBInsights, PitchBook and Tracxn data. Some round amounts are approximate where investors declined to disclose.
Why NVIDIA, Databricks and Bezos Matter
The Perplexity AI investors list is unusual because it combines AI infrastructure, data infrastructure and consumer internet capital. NVIDIA’s presence signals investor interest in high-frequency inference workloads, especially when AI search moves from occasional Q&A to persistent agentic retrieval. A search assistant that answers with citations must perform retrieval, ranking, synthesis and response generation repeatedly, which makes inference cost a core operating variable.
Databricks brings a different kind of signal. Perplexity’s enterprise story depends on structured and unstructured knowledge access, internal files and work apps. A data infrastructure investor suggests that the product’s long-term value may not sit only in consumer search. It may sit in the layer between organizational data, public web data and AI-generated answers.
Jeff Bezos’s involvement gives the investor list a consumer-internet angle. Search is a distribution business, not only a model-quality race. Reuters reported that Perplexity made a $34.5 billion unsolicited all-cash offer for Google’s Chrome browser in August 2025—a move that revealed how aggressively the company sees browser distribution as part of the AI search stack.
Product Research File: What the Investors Are Funding
Perplexity’s product is best understood as a retrieval-first AI interface. The consumer answer engine provides real-time answers with citations. The paid Pro plan adds higher model access, deeper research and file capabilities. The Max tier adds much heavier usage, Model Council (which routes a query through three frontier models simultaneously) and Perplexity Computer credits. The enterprise plan adds data privacy, admin features, team knowledge search and work-app connectors.
Perplexity’s developer layer is built around Sonar API, Sonar Pro, Sonar Deep Research, the Search API and the Agent/Agentic Research API. The Sonar API provides web-grounded AI responses with streaming, tools, search options and support for OpenAI-compatible client libraries. The Search API provides real-time ranked web results from a refreshed index, with filtering by domain, language and region. The Agentic Research API routes to third-party frontier models at first-party rates.
In our hands-on testing framework for editorial and B2B workflows, the practical value is strongest when Perplexity is used for source discovery, market scans, investor mapping, live company monitoring and citation-backed first drafts. The weakness appears when teams expect it to behave like a fully integrated internal knowledge platform out of the box. For Slack, Notion, Zendesk or deep proprietary knowledge-base workflows, buyers should assume middleware, API work or enterprise configuration will be required.
Feature, technical specification and API matrix
| Layer | Capability | Technical Detail | Buyer Relevance |
| Consumer | AI answer engine | Real-time answers with inline citations | Fast research & source discovery |
| Pro ($17/mo annual) | Advanced answers | Latest AI models; deeper research; file uploads | Knowledge-worker upgrade |
| Max ($167/mo annual) | High-usage workflows | 10,000 Computer credits; Model Council; advanced reasoning | Heavy research & agentic work |
| Enterprise ($34/seat/mo annual) | Team search layer | Web + team files + work apps; no training on customer data | B2B deployment |
| Enterprise Max ($325/seat/mo) | Regulated-org tier | Unlimited research; audit logs; SCIM; frontier models | Compliance-heavy teams |
| Sonar API | Grounded generation | $1/$1 per 1M tokens; 127K context; streaming; OAI-compatible | App integration |
| Sonar Pro API | Deep search model | $3/$15 per 1M tokens; 200K context; 2× citations; F-score 0.858 | Complex multi-source queries |
| Sonar Reasoning Pro | Multi-step reasoning | $2/$8 per 1M tokens | Agentic research pipelines |
| Sonar Deep Research | Autonomous investigation | $2/$8 base + citation + reasoning tokens + $5/1K search queries | Diligence & long-form reports |
| Search API | Raw ranked retrieval | $5 per 1,000 requests; domain / language / region filters | Raw search-result pipelines |
| Agent / Agentic Research API | Third-party model routing | GPT-5.x, Claude Opus 4.8 at first-party rates; $0.005/search call | Structured outputs & agents |
| Data privacy | Enterprise policy | No training on customer data; contractual guarantee | Legal & procurement review |
Complete 2026 Commercial Pricing Matrix, Including Hidden Limits
Perplexity’s current public pricing shows a clearer consumer plan structure than enterprise limit structure. Perplexity’s pricing page lists Pro at $20 per month ($17 when billed annually) and Max at $200 per month ($167 annually). Enterprise pricing starts at $40 per seat per month ($34 annually), with Enterprise Max at $325 per seat per month for regulated or high-volume deployments. Sonar API docs confirm token-based pricing from $1/$1 per million tokens on base Sonar to $3/$15 on Sonar Pro. The practical hidden limit is not only price—it is workload predictability.
| Plan | Public Price | Core Capabilities | Hidden Caps / Constraints |
| Free | $0 | Core AI search, citations, basic answers | ~5 Pro Searches/day; no frontier models; no API |
| Pro | $20/mo or $17/mo (annual) | Latest AI models; deeper research; 50-file upload | 20 research queries/day; 50 files at 50 MB each; no Enterprise SSO |
| Max | $200/mo or $167/mo (annual) | Model Council; 10,000 Computer credits; Sora 2 Pro video | Credit-based — unpredictable spend; 3 videos/mo without audio add-on |
| Education Pro | $10/mo | Pro features at 50% discount | SheerID verification required; no Enterprise features |
| Enterprise Pro | $40/seat/mo ($400/seat annual) | SSO, SCIM, SOC 2, 500 research/day, 15,000 files | ~50-seat minimum; advanced models excluded; audit logs limited |
| Enterprise Max | $325/seat/mo ($3,250/seat annual) | Unlimited research; full audit logs; frontier models incl. o3-pro | No seat minimum; highest per-seat cost; requires sales engagement |
| Sonar API | $1 input / $1 output per 1M tokens | Grounded Q&A, citations, 127K context | Per-request fees $5–$12/1K add on top of token cost |
| Sonar Pro API | $3 input / $15 output per 1M tokens | 200K context, 2× citations, best factuality score | Per-request fees $6–$14/1K; cost variance on long outputs |
| Sonar Deep Research | ~$2/$8 base + extras | Autonomous multi-step research | $0.30–$1.30+ per query; model controls search count |
| API rate limits | Tier 0–5 by credit spend | $50 to $5,000 lifetime credit tiers | 20–100 RPM ceiling; Tier 0 (no purchase) severely throttled |
Annual billing provides roughly 15–20% discount versus monthly. Enterprise Pro has a reported ~50-seat minimum; Enterprise Max removes that minimum, which can make Max the cheaper entry point for small regulated teams needing SCIM and full audit logging.
Hidden caps that catch B2B buyers
- Pro files: 50 files per Space at 50 MB each — heavy document workflows hit this fast and require an Enterprise upgrade.
- Research queries: 20/day on Pro versus 500/day on Enterprise Pro; unlimited only on Enterprise Max.
- Video generation: 3/month (Pro), 5/month (Enterprise Pro), 15/month with audio (Enterprise Max) — no à-la-carte top-ups.
- Frontier models (o3-pro, Opus-class thinking variants) are exclusive to Enterprise Max.
- Sonar Deep Research cost is variable: the model decides how many searches to run, so a single query can range from ~$0.30 to $1.30+ depending on reasoning depth.
- API rate limits ladder by lifetime credit purchase (Tier 0 no purchase to Tier 5 at $5,000), capping throughput at 20–100 RPM.
Perplexity AI Investors List: What the Cap Table Suggests
The cap table suggests three strategic bets. First, investors believe generative answer engines can capture high-intent search behavior. Second, they believe citations and live retrieval can differentiate Perplexity from closed chatbot responses. Third, they believe distribution can move beyond a website into browser, app, enterprise and API surfaces.
The risk is equally clear. Google can bundle AI answers into default search, Chrome, Android and Workspace. OpenAI can use ChatGPT as a general interface for research and agents. Anthropic can compete for enterprise reasoning workflows. Perplexity’s investor case therefore depends on focus: better answer UX, faster source grounding, stronger publisher relationships and a product architecture that keeps retrieval central.
A 2028 IPO target—confirmed by Reuters in June 2026—raises the bar significantly. Public-market investors will ask harder questions than venture funds: gross margin, net revenue retention, paid conversion rates, enterprise expansion velocity and legal exposure on copyright and publisher relationships.
Aravind Srinivas stated Perplexity’s 2028 IPO plan remains in place, acknowledging that failed AI listings could create broader ripple effects for the industry. — Aravind Srinivas, CEO — via CNBC / Business Insider (June 2026)
Implementation Workflow for B2B Teams
A practical Perplexity implementation starts with use-case separation. Teams should not begin with ‘deploy Perplexity everywhere.’ They should classify workflows into public web research, internal knowledge search, executive briefing, market monitoring, support enablement and API-powered product features. Public web research can be handled with Pro or Max seats. Internal work requires Enterprise evaluation. Application features require Sonar API, Search API or Agent API.
The second step is data policy review. Perplexity’s enterprise page emphasizes no training on customer data, but procurement teams still need contractual language, audit rights and administrative controls — especially for regulated industries where HIPAA eligibility and configurable retention are requirements.
Step-by-step technical deployment
- Define the query class. Separate simple lookup, competitive intelligence, internal knowledge retrieval, executive briefing and customer-facing answer generation.
- Select the product layer. Use Pro for individual researchers, Max for high-volume analysts, Enterprise Pro for teams, Enterprise Max for regulated orgs, and API products for embedded search.
- Configure data governance. Confirm no-training terms, retention policies, workspace controls, user access, connector permissions and acceptable-use boundaries.
- Choose the API. Use Search API for raw ranked web results, Sonar for fast grounded answers, Sonar Pro for deeper multi-source queries and Agent API for structured outputs or broader agentic workflows.
- Build request templates. Standardize system prompts, domain filters, source filters, region settings, JSON schema requirements and citation expectations.
- Set search_context_size. Sonar API requests accept Low, Medium or High context — higher context improves grounding but raises cost per request independently of the model’s context window ceiling.
- Add evaluation. Measure citation precision, answer correctness, latency, refusal handling, source freshness and cost per successful answer.
- Monitor bottlenecks. Track token volume, output length, citation load, repeated searches, context overflow and hallucinated or weakly supported claims.
- Create human review tiers. Finance, legal, medical, enterprise sales and investor relations workflows should require source verification before publication or decision use.
In a Business Insider report on Perplexity fundraising, Srinivas noted that the Series A was the only time he made a pitch deck, reflecting how the company uses its own AI workflow in investor communications. — Aravind Srinivas, CEO — LinkedIn / Business Insider (2025)
Known User Constraints and Performance Bottlenecks
The first bottleneck is source reliability. Perplexity’s value depends on live retrieval, but live retrieval inherits the web’s uneven quality. Academic research on generative search engines has found that citation-backed answers can still contain unsupported statements — making verifiability an ongoing challenge rather than a solved feature. Stanford researchers Liu, Zhang and Liang argued that generative search systems should cite comprehensively and accurately, a standard that directly frames Perplexity’s core trust proposition.
The second bottleneck is cost. API workloads can grow faster than expected when users ask multi-step questions, request long reports or trigger repeated searches. Sonar Pro is priced at $15 per million output tokens — meaningful at scale. For enterprise dashboards with thousands of users, the cost center is not a single answer. It is repeated retrieval plus synthesis across many sessions, compounded by citation tokens and reasoning tokens in Deep Research mode.
The third bottleneck is integration depth. Perplexity’s enterprise plan references team files and work apps, but buyers should test their actual stack. Slack, Notion, Google Drive, Zendesk, Salesforce, Confluence and internal databases each bring access-control and freshness problems. Without strong permissions mapping, an AI search layer can either miss critical information or expose it too broadly. The standalone architecture means custom middleware is the current norm for complex internal stacks.
Stanford researchers Nelson F. Liu, Tianyi Zhang and Percy Liang found that generative search systems should cite comprehensively and accurately — a standard that directly applies to Perplexity’s core product promise. — Liu, Zhang & Liang — Evaluating Verifiability in Generative Search Engines, arXiv 2023
Competitive Position Against Google, ChatGPT and Enterprise Search
Perplexity’s investor case depends on a narrow but powerful wedge: answer-first search with visible sources. Google owns default search distribution. OpenAI owns the largest chatbot mindshare. Anthropic has strong enterprise reasoning momentum. Perplexity’s differentiator is a product experience built around real-time research rather than open-ended conversation.
For enterprises, Perplexity’s closest substitute is not always Google. It may be a stack of tools: ChatGPT Enterprise for reasoning, Google Workspace for document retrieval, Glean for enterprise search, Microsoft Copilot for Microsoft 365 and custom retrieval-augmented generation built on internal data. Perplexity wins when teams need fast, source-linked research across the public web and selected internal knowledge, and lose patience with multi-tool context-switching.
What Most Investor Lists Miss
Most investor lists stop at names. The more useful analysis asks why those names cluster around Perplexity. NEA and IVP validate the venture pattern. NVIDIA validates inference economics. Databricks validates data architecture. Bezos validates consumer distribution instincts. SoftBank and Accel validate growth-stage appetite. Bessemer validates enterprise software relevance.
The second overlooked point is that Perplexity’s product and cap table are tightly coupled. If AI search becomes a high-volume consumer habit, the company needs capital for compute, distribution, publisher partnerships and model access. If it becomes an enterprise research layer, it needs security, permissions, connectors, compliance and sales capacity. If it becomes an API platform, it needs developer reliability, pricing clarity, latency management and benchmark transparency.
The third overlooked point is IPO timing. A 2028 target leaves roughly two years to prove durable revenue, improve unit economics and defend differentiation. Management is reportedly targeting $656 million in revenue by end of 2026, up from approximately $200 million ARR in late 2025. The investor list is impressive, but public-market investors will ask about gross margin, retention, paid conversion and legal exposure on copyright relationships with publishers.
Key Takeaways
- The verified Perplexity AI investors list includes NEA, IVP, NVIDIA, Databricks, Bessemer Venture Partners, SoftBank Vision Fund 2, Accel, Jeff Bezos and early angels including Elad Gil, Nat Friedman, Bob Muglia, Paul Buchheit, Tobias Lütke and Dylan Field.
- The strongest confirmed funding milestone is the January 2024 IVP-led $73.6M round at ~$520M valuation; later figures from $9B to $22.6B are widely reported but carry varying levels of confirmation.
- The product strategy spans Pro, Max, Enterprise Pro, Enterprise Max, Sonar API, Sonar Pro, Sonar Deep Research, Search API and Agent API — no longer simply a consumer search tool.
- Pricing is clearest for individuals: Pro at $20/mo ($17 annual), Max at $200/mo ($167 annual). Enterprise pricing requires sales engagement and carries hidden caps on research queries, file limits, video generation and frontier-model access.
- The biggest implementation risks are citation reliability, API cost variance (especially Deep Research), connector limitations, access-control mapping and source freshness.
- NVIDIA and Databricks are strategically important because Perplexity’s future depends on inference economics and data retrieval architecture — not just model quality.
- A 2028 IPO target raises the bar: durable revenue, gross margin and enterprise expansion velocity will matter more than headline valuation momentum.
Conclusion
The Perplexity AI investors list shows how quickly AI search moved from product experiment to capital-intensive platform battle. Perplexity’s backers represent nearly every layer of the modern AI economy: venture funds, chip infrastructure, data platforms, consumer internet operators and growth capital. That mix gives the company credibility, but it also clarifies the pressure it faces.
For B2B buyers, the investor list should be read alongside product architecture. Perplexity is strongest when the job is live, citation-backed research across the web and selected knowledge sources. It is weaker when buyers assume it will automatically replace a deeply integrated enterprise search stack without configuration, governance and evaluation.
Funding can buy time, compute and distribution experiments. It cannot by itself guarantee citation quality, enterprise trust, profitable API usage or durable search behavior. If Perplexity can turn its answer engine into a reliable operating layer for research, its investor base will look prescient. If search defaults remain locked inside browsers, operating systems and office suites, the cap table will look like a high-conviction bet against some of the strongest distribution machines in technology.
Frequently Asked Questions
Who are the main investors in Perplexity AI?
The main named investors include NEA, IVP, NVIDIA, Databricks, Bessemer Venture Partners, Jeff Bezos, SoftBank Vision Fund 2, Accel and early individual backers such as Elad Gil, Nat Friedman, Bob Muglia, Paul Buchheit, Tobias Lütke and Dylan Field. Tracxn counts 62 investors in total across roughly eleven rounds.
Which investor led Perplexity’s 2024 Series B?
IVP led Perplexity’s January 2024 $73.6 million Series B. Reuters reported the round valued Perplexity at about $520 million and included NVIDIA, Jeff Bezos, NEA, Databricks and Bessemer Venture Partners.
What is Perplexity AI’s latest reported valuation?
Reuters reported talks in March 2025 to raise capital at an $18 billion valuation. A September 2025 round was reported at $20 billion by Reuters and The Information. Early-2026 secondary-market and Tracxn estimates place the figure at $21.2–22.6 billion. Treat post-$9B figures as reported valuations unless Perplexity directly confirms a round.
What does Perplexity sell to businesses?
Perplexity sells Enterprise Pro ($40/seat/mo) and Enterprise Max ($325/seat/mo) for team search across web, files and work apps. It also offers APIs including Sonar API for grounded answers, Search API for ranked web results and Agent API for structured or agentic workflows.
What are the biggest risks for enterprise buyers?
The biggest risks are citation accuracy, source quality, API cost variance (especially in Deep Research mode), connector limitations, access-control mapping, rate-limit ceilings and overreliance on AI-generated summaries without human verification.
References
Business Insider. (2026, June). Perplexity’s CEO says it’s still aiming for a 2028 IPO, regardless of how OpenAI and Anthropic fare. https://www.businessinsider.com/perplexity-ai-ipo-plans-openai-anthropic-spacex-market-valuations-2026-6
Liu, N. F., Zhang, T., & Liang, P. (2023). Evaluating verifiability in generative search engines. arXiv. https://arxiv.org/abs/2304.09848
NEA. (2023, March 28). Our investment in Perplexity AI: Answer engines and the end of traditional search. https://www.nea.com/blog/our-investment-in-perplexity-ai-answer-engines-and-the-end-of-traditional-search
Perplexity AI. (2026). Enterprise pricing. https://www.perplexity.ai/enterprise/pricing | API pricing docs. https://docs.perplexity.ai/docs/getting-started/pricing
Reuters. (2024, January 4). Search startup Perplexity AI valued at $520 mln in funding from Bezos, Nvidia. https://www.reuters.com/technology/perplexity-ai-valued-520-mln-funding-bezos-nvidia-2024-01-04/
Reuters. (2024, November 6). Perplexity raising new funds at $9 bln valuation, source says. https://www.reuters.com/technology/artificial-intelligence/perplexity-raising-new-funds-9-bln-valuation-source-says-2024-11-06/
Reuters. (2025, March 20). Perplexity AI in talks to raise funds at $18 billion valuation, source says. https://www.reuters.com/technology/artificial-intelligence/perplexity-ai-talks-raise-funds-18-billion-valuation-bloomberg-news-reports-2025-03-20/
Reuters. (2026, June 9). Perplexity plans 2028 IPO regardless of Anthropic or OpenAI listings, CEO says. https://www.reuters.com/business/perplexity-planning-ipo-2028-regardless-what-happens-anthropic-or-openai-ceo-2026-06-09/
