What people are saying about Anthropic.

Gavin Baker’s initial reaction was not “Anthropic is worth $2T,” but rather: This $2T figure may be market speculation generated by competition among investment banks. His reasoning is very investment-banking-oriented: multiple banks are competing for the lead underwriting role. If one loses the mandate, it has an incentive to leak an extremely high valuation expectation to the media, making it harder for the winning bank to price the deal later. So he draws a distinction: “Anthropic may eventually trade at $2T-$4T” ≠ “The IPO should be priced directly at $2T.” He believes the IPO should leave some upside because Anthropic’s most important asset is its employees. If the IPO is priced too high, post-listing volatility and a decline after lock-up expiry could seriously affect employee morale and retention. This is actually a very mature assessment: do not treat a valuation target reported by the media as the company’s intrinsic valuation.

▶

YouTubeWatch on YouTube

I think LLM is too competitive. But Anthropic is good because of it's strong ethics

I like that Claude gives thoughtful, well-structured responses even to complex prompts.

I know almost nothing about programming, but with Claude, I wrote over 200 lines of code that can browse the web efficiently.

I had a very good first impression of Claude, but after subscribing, I quickly hit the message limit, especially when an issue couldn't be resolved within just a few exchanges. This was very frustrating.

My take: revenue is growing fast, but so is the burn.

I think $965B already prices in flawless execution, on top of $180B+ in compute commitments and a Pentagon supply-chain risk designation no other tech company has ever gotten, and that combination worries me.

My own opinion, if gross margins hit 77% and revenue reaches $70B by 2028, the long-term bull case easily clears $1 trillion-plus.

I think Amazon and Google stand to benefit threefold from this round: they are cloud infrastructure providers, chip suppliers, and shareholders. As Anthropic's valuation rises, all three lines benefit.

With $47B run-rate revenue and the margin profile we underwrote, the question isn't whether they hit a trillion, it's whether the mark is conservative.

In my due diligence I found 8 of the Fortune 10 are Claude customers, and Claude Code alone is generating $2.5B in annualized billings. I haven't seen enterprise penetration move this fast anywhere else.

What I’m most dissatisfied with are the message/conversation limits, the lack of shared context across different products, and the relatively weak image generation capabilities.

As a technical founder, Claude Code has the best "vibe" for me right now. Anthropic keeps polishing real workflows instead of flashy demos.

I like that Claude's answers are clear and keep context in long chats — though sometimes I find it overly cautious and verbose.

I use Claude 3.5 Sonnet for LeetCode problems, and it feels like it completely blows GPT-5.2 out of the water.

While GPT models are sometimes smarter, I really don't like their style. Claude just writes much more pleasantly.

I prefer Claude for coding tasks — its huge context window and Artifacts feature really help.

The world's most highly valued large language model company

Claude is my favorite AI tool.