Why is everyone always using UI to benchmark. Let them build a complex k8s stack with terraform etc....
2026-09-30 05:37:25
16
SQUARE :
Bro I only use Luna that’s my best friend when it comes to cost management and long sessions coding 🥺
2026-09-30 15:51:49
0
ctrl_c :
6.1 SOL is not as good as Astra. It’s definitely not. I was testing it last night. It’s better than 5.6 SOL. It’s looser then Astra, with less independent focus and even worse language description output.
2026-09-30 07:42:16
1
RedShower :
is that why i bécane 65000 crédit this morning?
2026-09-30 07:15:25
3
TikTokisGoogle :
Great comparisons. Next time put tokens to $$$ because sometimes we scroll and don’t catch current token costs.
2026-09-30 06:05:19
2
killebrewpatrick :
Astra felt like it was awesome when I first started using it. I've been through two week long projects where Astra invented it's own way instead of following the guide and roadmap that I gave it. Fable 5.1 had to rescue both of those software projects. Astra needs to be watched like a hawk on long term work for alignment with your expectations. Beware.
2026-09-30 14:13:56
0
nimbic :
Besides how much it burns through usage... how do you figure Astra was underwhelming?!?? I have found it to be an absolute game changer! It's one of the most meaningful releases they've ever had. I can now ask codex to build 3D models and animate them and walk away. My wife had it create presets for her in photoshop, it automated the entire process. It was a massive leap in productivity
2026-09-30 08:04:37
3
John Weagle :
idc how much better sol is than sonnet they screwed over customers
2026-09-30 07:59:16
3
ahdam,samad :
d-day v2
2026-09-30 06:54:28
1
אלמוג בונה עם AI :
Great video! Amazing work
2026-09-30 15:12:58
0
longsmithstuffing :
You think 6 was underwhelming 😂
2026-09-30 14:50:05
0
tacotruckguy44 :
Sonnet 4.5 was goated. Just saying
2026-09-30 09:17:50
0
LK :
it's bad
2026-09-30 12:28:03
0
Aleks Sense :
To eat your credits
2026-09-30 10:56:18
0
tdfc :
All of OpenAI models are just so slow on tokens per second… I refuse to pay the crazy ultrafast tax
2026-09-30 09:14:10
0
Stevebidi :
Benchmark is rubbish. Astral is still beating Sol 6.1 in my environment for real implementations
2026-09-30 14:19:14
0
1776Breaks :
@1776Breaks:so as a loyal pro user... you cut my useage in half for the same amount 20x down to 10x. THEN you charge $300 more per month for a 25% of from what i was getting.... then you offer a $2500 credit which is not damage control it is wrap up your projects on us as you exit. Inference is expensive pro is their heaviest users. they are politely telling us to exit. i cant imagine this lricibg and these changes being ok with anyone....
2026-09-30 11:12:45
1
To see more videos from user @chase_ai_, please go to the Tikwm
homepage.