Milan

[Milan @cosmic_spec]

@jeremyphoward @Zai_org I will take your word for it, and incorporate it in my workflows

[Milan @cosmic_spec]

@theo Its def step function improvement over opus. Have been using it on my codebases since release

[Milan @cosmic_spec]

@analogalok Model is bad, based on my few initial tests

[Milan @cosmic_spec]

@aravindr93 Plans for open weights?

[Milan @cosmic_spec]

@TeksEdge @MSFTResearch link?

[Milan @cosmic_spec]

@himanshutwtxs @theomkute @LocalDev_ Codex

[Milan @cosmic_spec]

@picoito I don't you at all, but congratulations 👏

[Milan @cosmic_spec]

Qwen3.6 27B > Minimax M 2.5

[Milan @cosmic_spec]

@0xSero Qwen 3.6 vs Gemma 4 31B, thoughts?

[Milan @cosmic_spec]

@LottoLabs @ThePrimeagen Comparable to say haiku 4.5?

[Milan @cosmic_spec]

@0xSero Everyone's talking about q3.5 27b model, if you have used to enough, what you can compare it to, haiku 4.5?

[Milan @cosmic_spec]

@dotta paperclip doesnt support hermes agent though right

[Milan @cosmic_spec]

@LottoLabs Qwen 27b with Hermes agent, how good at a high level? Usable?

[Milan @cosmic_spec]

@HarveenChadha Better name it, SharmajiKaLadkaGPT

[Milan @cosmic_spec]

@HarveenChadha Way low, the memory bandwidth

[Milan @cosmic_spec]

@ivanfioravanti Any eta on m5 ultra studio?

[Milan @cosmic_spec]

@HarveenChadha Congratulations sardarji, very happy for you and sarvam.

[Milan @cosmic_spec]

@DorianDevelops Kimi 2.5

[Milan @cosmic_spec]

@ashen_one 512 M5 ultra would be able to run minimax and glm5 quant, so one should go and buy 512 memory for sure

[Milan @cosmic_spec]

@Dating_Dynamics Its a simple fact, coz of olfactory adaptation