# Anthropic IPO draws mixed market response as Google and custom AI chips come into focus

By Riza Dagoc

Canonical URL: https://www.tokenpost.com/news/insights/62400
Published: 2026-11-21T22:40:23.000Z
Updated: 2026-11-21T22:40:23.000Z
Section: Insights & Views

> Ejaaz called Anthropic's early trading reasonable as investors assess its IPO. He also highlighted Google's AI models and the push by leading labs to develop custom chips.

Investors are weighing how to value Anthropic following its public listing, while competition among leading AI companies is extending to models and custom chips.

In a Yahoo Finance interview on Nov. 19, AI podcast host Ejaaz said Anthropic's early share price was relatively reasonable and that the market was still deciding how to assess an IPO of its size.

Ejaaz said there was evidence of hundreds of billions of dollars in buying demand before the listing. The stock rose about 34% above its offering price before falling to around $95 within roughly two hours, putting the company's valuation at about $1.7 trillion. He said that was a reasonable price from a balanced perspective.

Both bulls and bears had expected more extreme price swings, Ejaaz said, suggesting investors were uncertain how to value such a large IPO.

He also pointed to a rocky period for stocks over the previous four to eight weeks, including declines in semiconductor shares and a large drop in Nvidia's market value. Investors were considering how to compare Anthropic, which is highly exposed to the AI industry, with Meta, Google and Amazon, which also have substantial businesses and revenue outside AI, he said.

Ejaaz offered a positive assessment of Google's AI models. After being bearish on them for much of the year, he said he had tested three to four major models released over the previous three weeks and found them competitive. He said Google's models compared well with those from Anthropic and OpenAI while being cheaper and, in several respects, more capable.

He also said Microsoft could catch up in developing a leading AI model. Its access to OpenAI model intellectual property and products used by companies around the world give it advantages, he said, even though its own model development has yet to produce comparable results.

Ejaaz expects custom AI chip development to spread across leading labs. Chips designed for a company's own models could deliver more performance from the same computing resources, reducing costs and improving revenue, he said. He added that Microsoft has yet to build a strong model, making it uncertain whether the company could gain the same benefits.


