Uttkarsh

[Uttkarsh @uttkarsh26]

@dalethomas50 @Teknium It feels like
Glm5.2 gets the intent
While k2.7 code gets the instructions better

[Uttkarsh @uttkarsh26]

@dalethomas50 @Teknium I ran some benches in hermes using gpt 5.5 imitating my actual scenarios and skills

Using neuralwatt

K2.7 code beats GLM5.2
33-29 out of 39 points total

Not sure if this is correct

[Uttkarsh @uttkarsh26]

Xiaomi mimo referral code

I'm on Xiaomi MiMo Open Platform — running Xiaomi's flagship MiMo V2.5 and the rest of the lineup. Sign up with my code and you'll instantly get $2 in API credits. Code: UDPC9J · https://t.co/DPBzQP8jT0 (After signup, enter the code at the bottom-left

[Uttkarsh @uttkarsh26]

@KESCoHQ why no electricity in kblock lic

[Uttkarsh @uttkarsh26]

@lemonkillergame @ollama Definitely usage is more but ollama cloud models feel dumber than opencode go version of the same model or even crofai for that matter.
Anyone else feels the same ?

[Uttkarsh @uttkarsh26]

@NousResearch Why is telegram model picker just a text blob now instead of interactive selection buttons
Talking abt /model command via telegram

[Uttkarsh @uttkarsh26]

@KESCoHQ voltage fluctuating in lic k block yet again
Last time was phase jumper fault
Please get it checked asap

[Uttkarsh @uttkarsh26]

@kimmonismus Quantized version? Glm5.1?

[Uttkarsh @uttkarsh26]

@neural_avb @thdxr @Kimi_Moonshot Is there it ollama issue?

https://t.co/CDNOp8W3E3

[Uttkarsh @uttkarsh26]

completely unusable today @ollama #ollama

[Uttkarsh @uttkarsh26]

@stevibe Bt why ? Isnt glm5.1 and k2.6 smarter
Although slow on ollama most of the times and sometimes pathetic availability

[Uttkarsh @uttkarsh26]

completely unusable today @ollama #ollama

[Uttkarsh @uttkarsh26]

@ollama why we ahve this today? happened two-three times already, in opencode

[Uttkarsh @uttkarsh26]

Also has no one experienced excessive thinking evenf or simple tasks by kimik2.6 cloud via @ollama on @opencode ?

[Uttkarsh @uttkarsh26]

@ollama @FreedomGoat_ @smdcapital @Kimi_Moonshot Also one thing I have noticed
In opencode speed is fine but with hermes it’s not? Am I doing something wrong?

[Uttkarsh @uttkarsh26]

@ollama @FreedomGoat_ @smdcapital @Kimi_Moonshot I understand but unusable and too slow to get any meaningful work done with k2.6 and glm5.1 both
Emailed

[Uttkarsh @uttkarsh26]

@FreedomGoat_ @smdcapital @ollama @Kimi_Moonshot Same feeling
Will i get a refund?

[Uttkarsh @uttkarsh26]

@nahcrof Doesn’t have vision in opencode?

[Uttkarsh @uttkarsh26]

@neural_avb @Naman_Arora7 People saying too slow with glm5.1
Is is the case for you as well and wats ur experience with k2.6 bro?

[Uttkarsh @uttkarsh26]

@neural_avb @Naman_Arora7 Which plan/tier?

[Uttkarsh @uttkarsh26]

@AlexFinn To be honest props to them but in real world I bet even somnet performs better in complex scenarios
Heard this too many times now