Topic
Public listings by artificial-intelligence developers and the pricing, lockup and aftermarket dynamics that follow them.
New Constructs puts Anthropic's equity at as little as $144bn, 93% below the $2tn it is reportedly targeting for its IPO. That is its lowest scenario, built from reports of leaked financials because the prospectus was filed confidentially.
Reality
OpenAI roughly matched Anthropic in September on spending by about 120,000 shared customers, up from a quarter at the start of the year, OpenRouter data shows. Both labs are now cutting small-model prices before planned IPOs, so share costs each of them revenue per request.
Reality
Anthropic launched Claude Haiku 5.5 at $0.10 per million input tokens, 75% below Haiku 4.5, ahead of a planned IPO. Yahoo Finance reported the rate matches OpenAI's smallest model, giving buyers a matched quote before anyone can see what the cut costs Anthropic in margin.
Perspective Coverage
8 publishersReality
Anthropic aims to start marketing its IPO by the week of Nov. 9, according to Bloomberg, with investors putting its value near $2 trillion. Teams building on Claude will soon be able to read the vendor's losses and spending plans in quarterly public filings.
Perspective Coverage
5 publishersReality
CFO Sarah Friar told staff OpenAI "will be a public company in 2027," sooner if growth holds. The date is also a deadline on every private round and buyback between now and then.
Perspective Coverage
4 publishersReality
A cybersecurity sales operator now owns the number OpenAI's 2027 listing will be judged on, and Anthropic may reach the public market first.
Reality
Second-quarter revenue of $11.6 billion put Anthropic ahead of OpenAI for the first time. The IPO story that follows is about who can bill, not who benchmarks best.
Perspective Coverage
3 publishersReality
Alex Karp told CNBC that frontier labs are angling for government liability cover because their commercial clients would otherwise sue them. Buyers whose data feeds those models should read their contracts with that in mind.
Reality
David Sacks wants the listing paused until a departing researcher's safety claims are checked, which turns a governance argument into a calendar problem for a deal that has to price before the November midterms.
Reality
What Anthropic wanted from Decart was cheaper serving, not a new product line, and the walk-away leaves the buy-versus-license question on compute efficiency sitting exactly where operators keep finding it.
Reality
Ben Mann's Labs group shipped Claude Code at a hit rate he puts at 20 to 30 percent, and the cheap part of that machine to copy is the two-week kill review rather than the research signal feeding it.
Reality