Hacker News Reader

Comment by qwerasdf5 | Hacker News Reader

For those that care about the truth: This is a non-story.

The $70b estimate was based on a comparison to Anthropic, which includes revenue from cloud providers. OpenAI does not include this in their numbers.

So, the number did not come from OpenAI. It was an accounting mistake made by some investors and media, who did not adjust for this before reporting on it. I see fintwit calling on FT/Axios to issue a retraction; good luck with that.

It sure is interesting to see the rush to judgement in this thread. Another poster correctly pointed out this mistake (now buried under the sea of misinformed posts trending above it.) The Internet is cooked.

View on HN

"The $70b estimate was based on a comparison to Anthropic" -- Can you prove that? All I could find were major outlets (Reuters, Yahoo Finance) citing an Axios report, that itself cited "sources familiar with the financials tell Axios.". I haven't found anything similar to what you are claiming.

Well, Axios are now reporting this themselves: https://www.axios.com/2026/10/08/openai-50-billion-arr-anthr...

Also bear in mind that the original $70b number came from an 'anonymous' source to Axios.

It says the 70B figure "was based on _information shared with investors_ in an attempt to create a more direct comparison to rival Anthropic".

That's a different thing from what you said, that the 70B "was based on a comparison to Anthropic". It wasn't, or we can't say that using the source material. The 70B was based on an anonymous source, which used the different metric (including cloud revenue etc) supposedly to make a comparison with anthropic, which does those things, easier.

Anyway, it doesn't seem like this is "not news"

> That's a different thing from what you said

Huh? Sorry, I don't see the confusion.

> Anyway, it doesn't seem like this is "not news".

Given that the number in the headline is based off of a nonsensical apples-to-oranges comparison of revenue figures, I think at a minimum we've got a misleading headline, don't you think?

Reply on news.ycombinator.com