@opencode did you take a picture of your computer screen? lmao

@opencode Is operation cheepseek still in motion?

@opencode heard great things about GLM 5.3 flash… ima try it myself tomorrow

@opencode hey, is GLM-5.3-flash meant to be available in Zen?

@opencode interesting asf

@opencode Deepseek bouncing back fast and Muse Spark is catching up

@opencode Hey what happened to performance. Mimo 2.5 is so slow and getting lots of errors.

@opencode I wonder why

@opencode But usage limits are being reduced for Go subscribers.

@opencode I'm finding that GLM 5.3 Flash is about as good cost wise relative to performance. Its' thinking traces are enormous, though.

@opencode @OverfitForTruth Many, like me, will have gone for the z.ai subscription rates.

@opencode It’s expected with this price , specially when it was FREE last week.

@opencode Yesss muse finally getting the respect it deserves

@opencode Speed matters for majority, and the value for money with DS flash cannot be compared with any models right now

@opencode It's all because you cut the quota way too low, so even a flash model can't work properly

@opencode now that all the freeloaders are gone, I can finally milk this cheap ass model without the entire world using it 🥰

@opencode Once you get GLM-5.3-Flash to the 60$ usage it will probably jump up.

@opencode benchmarks shift daily at this point

@opencode I noticed a discrepancy with the OpenCode Go usage limits.
The official docs state:
$12 / 5h
$60 / month
But my dashboard currently shows:
$6 / 5h
$30 / month
So it looks like all Go limits are currently 50% of the documented limits. !??

@opencode the award for most confusing sorting order goes to

@opencode Hy3 is very good despite tiny context window

@opencode You go out, you have a couple. You hook up with some stunner.

You wake up the next morning and realize it’s GLM 5.3 flash.

You sneak out. You don’t leave your number.

@opencode Don't remove muse spark contributer please

@opencode I've been quite impressed with Hy3, capable and pretty fast

@opencode when operation cheepseek continue?

@opencode GLM-5.3-flash is not Ox Alpha , I used it every day exclusively and as soon as it was revealed and“became paid”, the quality dropped to the point of being unusable

@opencode I don’t know but I feel Luna is perfectly fine , fast abd cheap for my use.

@opencode Will we ever get proper 60$ usage on models

@opencode Muse spark capabilities are... Questionable

@opencode Who knew that low cost llms would be popular

@opencode muse is like actual malware. i dont mean the data harvesting i mean the agent itself is so lobotomized and retarded i wouldnt let it update a readme

@opencode Why is this not sorted

@opencode Not a surprise, but I'm loving my GLM subscription even more!

@opencode who tf is using muse bruh