Hacker News Reader

Comment by eshack94 | Hacker News Reader

Asking the real questions. I've been wondering what the holdup on that is.

Does anyone reading this have additional knowledge or insight on this?

Sonnet, Opus, and Fable are pushing so much revenue growth right now that it makes more sense to keep growing the expensive models than growing the cheap models.

Yep. Simple answer is they want to IPO in the fall, and a new Haiku does literally nothing for them

They haven't really mentioned practically anything about Haiku in quite a while so I imagine nobody except for people inside Anthropic will have any indication.

Maybe it'll come out eventually but they don't even include it on some of their comparison benchmarks anymore, so I figure its very low priority for them.

More likely that they are embarrassed by how their attempts compare with OpenAI's Haiku analog, Luna.

Not so sure since Anthropic has 4 model families while OpenAI has 3 for GPT-5.6.

Claude Fable/Mythos vs GPT-5.6 Sol

Claude Opus vs GPT-5.6 Terra

Claude Sonnet vs GPT-5.6 Luna

Claude Haiku vs ?

The pricing on Luna is just insanely good. Haiku simply cannot compete at almost any intelligence level against DeepSeek V4 Flash or Luna.

No, you're off by one - likely because the OpenAI models punch above their weight in those comparison, hence my original message. You've shifted the comparisons to favor Anthropic.

Fable/Mythos are much larger than Sol. They match to Astra which is supposedly at least 10T. Astra is already publicly confirmed as a new family.

Opus matches to Sol. Sonnet to Terra. Haiku to Luna.

Anthropic is able to compete at the frontier high-end by launching massively large expensive models. But their inability to compete on small models belies their efficiency aspirations across the stack.

Reply on news.ycombinator.com