Perplexity eyes 2028 IPO despite Anthropic, OpenAI outcomes
AI
IPO
OpenAI
Anthropic
Perplexity
TLDR: SAN FRANCISCOâPerplexity CEO Aravind Srinivas told CNBC the company will pursue an IPO in 2028 regardless of Anthropic and OpenAI IPO outcomes. The timeline adds pressure as investors judge mega listings and AI frontier valuations.
Key Takeaways:
- Perplexity is positioning itself as an AI broker while Anthropic and OpenAI confidentially filed for IPOs and markets prepare for multiple mega listings.
- Srinivas said Perplexity is planning for 2028 IPO timing regardless of how Anthropic or a reported OpenAI listing lands, calling their models frontier work.
- Upcoming IPO performance could create ripple effects across AI valuations, pushing frontier labs to show steady model capability advances and to manage AI spend carefully.
- Srinivas also highlighted AI procurement pressure, saying Perplexity aims to choose the cheapest model that still hits task quality, using frontier tools only when needed.
The wild part is the confidence. Perplexity is treating the IPO window like a destination, not a popularity contest, while frontier labs race to prove their upgrades keep receipts.
The wild part is the confidence. Perplexity is treating the IPO window like a destination, not a popularity contest, while frontier labs race to prove their upgrades keep receipts.
Q&A
If Anthropic and OpenAI IPOs stumble, how might that change Perplexityâs pricing and investor base in 2028?
A weak debut could tighten valuation expectations for AI in general, pushing Perplexity toward more conservative pricing, steadier revenue narrative, and investors focused on profitability and unit economics.
What does âfrontierâ mean to investors when Srinivas ties valuation to six months of model progress?
It signals a quarterly reality check. If frontier labs slow upgrades for about half a year, investors may treat leadership as slipping rather than merely iterating.
Why is Perplexity talking about model selection and cost now, before its own IPO is near?
It reframes growth as operational efficiency, not just scale. Showing control over inference spend and task accuracy helps investors believe margins can survive competition and capex cycles.
Could âtokenmaxxingâ shift enterprise buying from raw usage to measured outcomes, and how might Perplexity benefit?
Yes. When buyers demand productivity evidence instead of usage counters, Perplexityâs model choosing based on cost and task fit can become a sellable advantage.
What happens to AI brokerage strategies if the frontier models become too expensive for most customers?
Brokerage becomes more valuable. If frontier quality stays high but costs spike, systems that route tasks to cheaper models that still hit quality thresholds can win more enterprise budget.
Top in Startups
Ferveretâs reactor inspired cooling pressures data centers to rethink efficiency
SpaceX IPO courts investors with AI and multiplanetary ambition
SpaceX IPO splits bull and bear bets ahead
OpenAI IPO filing turns AI funding race up fast
No comments yet. Be the first to share your thoughts!