Anthropic's IPO, and a $3 Trillion Valuation Case: All-In Podcast Reads the 'Big Two' Duopoly
TL;DR · What you'll learn
- 1 The All-In Podcast dedicates a segment to the 'trillion-dollar IPO rush' -- following SpaceX's listing, both OpenAI and Anthropic are being watched for a possible IPO by early next year.
- 2 Anthropic reportedly filed confidentially for an IPO on June 1st, and Polymarket is pricing a 65% chance the IPO happens this year.
- 3 Prominent investor Gavin Baker reportedly predicted Anthropic will hit over $100 billion in annual revenue with strong margins by end of 2026, and would trade at a $3 trillion valuation if it went public today.
- 4 Panelists noted OpenAI still burns significant cash due to its heavy reliance on consumer business, while Anthropic may be 'accidentally profitable.'
- 5 One panelist argued that switching to open-source models carries real technical friction that can't match the ease of simply 'firing up Claude.'
- 6 DoorDash's CTO was cited as having disclosed a hybrid setup: handing the hardest work to Anthropic's frontier model Fable, while routing lighter tasks to Kimi 2.6.
- 7 By token-revenue terms, Anthropic's ARR was pegged at roughly $60 billion versus OpenAI's roughly $40 billion, with the panel arguing AI is already consolidating into a 'top-two' duopoly.
Read as slides
3 slides total
A Trillion-Dollar IPO Rush -- Is Anthropic Next?
With SpaceX now public at a $2 trillion valuation and trading near its IPO price, the panel turned to OpenAI and Anthropic as the next candidates for a 'trillion-dollar IPO.' Anthropic reportedly filed confidentially on June 1st, with Polymarket pricing a 65% chance of an IPO this year.
Investor Gavin Baker's forecast -- that Anthropic could hit over $100 billion in revenue with strong margins by end of 2026, and would trade at a $3 trillion valuation if public today -- was cited. Panelists noted OpenAI still carries heavy cash burn tied to its consumer-heavy business, while Anthropic may be 'accidentally profitable.'
Meta's 'Price War' and DoorDash's Hybrid Setup
The conversation turned to Mark Zuckerberg's tweet declaring a price war -- offering 'the same quality at 1/100th the cost' -- as a sign that cost is becoming the new competitive axis. One panelist countered that switching to open-source carries real technical friction that can't match the simplicity of just firing up Claude.
As a real-world example, DoorDash CTO Andy Fang was cited as having disclosed a hybrid architecture: internal coding benchmarks let them route the hardest work to Anthropic's frontier model Fable, while delegating lighter tasks to Kimi 2.6.
The Revenue Numbers Behind the 'Duopoly'
By token-revenue terms, Anthropic's ARR was pegged at roughly $60 billion versus OpenAI's roughly $40 billion -- panelists argued the AI industry is consolidating into a 'top two, then everyone else' structure. A caveat was noted, though: open-source model usage counts as 'dark tokens' that never register as revenue, so this metric alone doesn't capture actual usage share.
Editor's Take
What this episode really shows is that frontier AI labs are increasingly being read through the lens of business metrics -- ARR, margins, IPO valuations -- rather than raw technical capability, a sign of how mature the industry has become. DoorDash's hybrid routing setup is the practically useful takeaway here: pairing a frontier model for hard judgment calls with a cheaper model for routine work is likely to become a standard architecture pattern, and Claude Code developers should watch this trend closely for cost-optimization purposes.
Source
All-In Podcast
OpenAI vs Anthropic IPOs, Anthropic $3T, Zuck's Price War, China Ends Open Source?, Trump Accounts
This article auto-summarizes the YouTube video's transcript with Claude. Please refer to the original video for nuance and exact wording.
Watch on YouTube →Related
3 articles
Brock Mesarich | AI for Non Techies
Anthropic Ships Claude Cowork Mobile: Tasks Keep Running Even After You Close Your Laptop
Anthropic released a mobile version of Claude Cowork, calling it one of the most requested features it has seen.
Inteligencia Artificial
Behind Fable 5's Comeback, Anthropic's Relationship With Washington Sours -- Could Europe Become a 'Safe Harbor'?
Fable 5, previously pulled offline, is back and currently the world's most powerful AI model -- included in standard plans until the 7th, after which usage-based billing returns and prices rise.
AI Engineer
An Anthropic Engineer's 'Field Guide to Fable': Closing the Gap Between the Map and the Territory
A member of the Claude Code team spoke at the AI Engineer conference and described Fable as a model where the open world suddenly opens up.