mydreamof

I see I am the only one that is actually surprised that LLM are so good on this versus tool that is specialized to do one thing

I realized that I am no more frustated that I have to solve conflict :D Did you notice the same? I remember that I had hadache just about thinking that I will need to solve it...

I don't think new benchmarks are saturated. They still give you a clue, they arn't perect but they have value. If model can't even do some easy tasks from benchmark then why would u even consider using it?

Cost are bigger than Opus 5.5 for that effort

bro agent scripts doesn't matter at all imo

In the other hand the pricies dropped by a big margin

It is a 9B agentic model developed by Xiaomi MiMo through supervised fine-tuning of Qwen3.5-9B on MiMo-generated data

Slower, less reliable but cost more :D Of course this is max reasoning so I guess it make no sense to use it

Seems like benchmaxing? For example for Terminal-Bench 4 it doesn't have great results. And why not show other benchmarks?

It is more token efficient I believe

Inspired by everyuuid.com

What I found interesting is that when you click a random line multiple times there is almost always plenty of queens. On averege there is 5.5 queens because the whole space is dominated by near full boards.

If you are interested more in technical aspect I recommed reading about everyuuid: https://eieio.games/blog/writing-down-every-uuid/

Ofcouse it will go down. Look at cost of Fable vs GPT-6 - it is already 2.3 cheaper

I don't get it. For me it seems Opus was more accurate in terms of for example this small building in the right down corner

Wow its better than Fable5 and GPT5.6 SOL in many benchmarks

Big jump in benchmarks

IMO it depends on the company. If everyone in the company has to provide a lot of PRS etc. you have to stop care about the quality and just get the paycheck. The CEO decided not you. (Of course not when the product may charm ppl etc.)

Bro these colors on chars are unbelievlable, I can not understand which is opus, which is fable, which is GPT...

you can even use gpt 1.0 but it proves nothing

there is no GPT 5.6 init, so what's the point?

It would be great to see some benchmark how it improve the agent. Like agent will be better at getting information from such data or what? Otherwise what is the goal of using it