Anthropic IPO - Hackdot

Re:Anthropic IPO by CamperBob2 on Thursday November 12, 2026 @11:48AM (#50470280)

So much of "dramatically ahead" is due to the Claude Code harness being better than anyone else's.

Fable is great, but it's no better at writing or reviewing code than the top-end Carbon or Astra models. To the extent Anthropic has a moat, it's Claude Code. Obviously that's not much of a moat, but their competitors seem content to camp on the other side of the drawbridge for some reason.

[ Parent | on HN ]

Re:Anthropic IPO by runtime_terror on Thursday November 12, 2026 @11:53AM (#50470350)

Open source ones like OpenChamber, etc are doing good work in competing on this front (including being able to use your subscription credits directly)

[ Parent | on HN ]

Re:Anthropic IPO by orbital-decay on Thursday November 12, 2026 @11:56AM (#50470403)

What makes it a moat? How did you compare? I don't think CC has particularly outstanding integration (used in practice) or performance. People use it because Anthropic shoves it down their throats, making it risky to use anything else with their subscriptions, and generally gatekeeps their models like crazy.

[ Parent | on HN ]

Re:Anthropic IPO by CamperBob2 on Thursday November 12, 2026 @12:06PM (#50470567)

I use all three major harnesses (Antigravity, Codex, and CC) and CC is better. It gets used for writing code, while the others are used only for reviewing what Claude writes.

Codex is pretty close in a lot of ways, as the other poster says, but Astra consumes tokens at a higher rate, and when it compacts it completely lobotomizes itself. That used to be a problem in Claude Code as well, but it has been largely solved (although at the cost of making the compaction take much longer.)

Antigravity reminds me of the old comic strip where Firefox and Chrome are duking it out, while Internet Explorer is sitting in the corner wearing a helmet and eating out of a jar of paste.

[ Parent | on HN ]

Re:Anthropic IPO by unpwn on Thursday November 12, 2026 @12:27PM (#50470893)

I use a 3P harness, and Claude performs significantly better than any of the other models

[ Parent | on HN ]

Re:Anthropic IPO by orbital-decay on Thursday November 12, 2026 @12:33PM (#50470977)

I don't see in your reply that you attempted to disentangle models from harnesses, it seems you're mixing them up. I regularly test different harnesses on different models and for my use cases CC is just about as performant with OMP on Claude, for example, but it takes a lot more tokens. There's nothing special about the harness itself.

[ Parent | on HN ]

Re:Anthropic IPO by CamperBob2 on Thursday November 12, 2026 @01:30PM (#50471907)

I don't see in your reply that you attempted to decouple models from harnesses, it seems you're mixing them up.

Normally I run the models that the harness is intended to support. You can't run Claude models anywhere else besides Anthropic's own CLI and desktop harnesses -- or at least you're not supposed to -- without paying out the nose for API tokens.

I do use DeepSeek 4.1 Flash 1017 in Claude Code when I want an abliterated model for security-related work or unhobbled browser control, though. I know it's possible to get OpenAI models working in CC, but I haven't bothered. Gemini I'm not sure about.

[ Parent | on HN ]

Re:Anthropic IPO by efromvt on Thursday November 12, 2026 @12:02PM (#50470490)

I mean, OpenAI has codex? I find it pretty equivalent. So they are trying to equalize. (Less said about Google CLI attempts the better). And there’s definitely not model equality - Astra was a big step forward but generally loses on daily utility to opus 5.5, much less fable. I would be very scared for Anthropic if it was trying to compete purely on harness; I suspect most of its usage is by default because the models are better. In the pre 5.5 phase I saw a lot of migration to codex/pi, and I’d be fine without CC as a harness today.

[ Parent | on HN ]