KaoruAoiShiho

You're probably just responding to the headline but this person is an AI bull and isn't claiming it's a big deal, she's going into it and explaining it.

Click through to the link, the answer is no it uses the latest gpt models now.

I think the parent's point is that if you are genuinely open to losing, the arguments can be productive because you can learn something instead... So stopping arguments is just another way of closing yourself off.

And before you know it, you invented some openrouter provider from first principles...

Are you sure fireworks is unquant? It's not listing precision on openrouter like everyone else.

This is really held back by one bench (omniscience accuracy) where it's really very far behind otherwise i think it's got at least a couple of points higher.

Fable largely fixed the annoying chatterness so sucks that it's gone now.

Then that's just a video game might as well as play a video game why limit yourself to still confined to the rules of chess.

Kimi 2.5 has the best long context. For raw coding benchmark scores you can just post train on top of it with more specialized data. 2.5 is kinda old, 2.6 is the current release which is exactly just that and catches up to the frontier in most aspects.

Do people like "gacha"? I thought people played games for the game experience, story, etc, and the gacha is just the monetization mechanic. It's like making a big deal out of paying $20 bucks a month, or buying loads of DLCs for example.

No I think the best agent with hundreds of millions in ARR should be worth more than the 15th best model company with tiny revenue. ur the joke.

Manus is saved, 2 billion is such an undervaluation considering much worse companies like minimax is valued at 30 billion.