7A7z

@Kappaemme1926 the day before yesterday x.com/7a7zz/status/2…

be honest when was the last time you wrote a line of code?

@sadkatwt @Pirat_Nation where tf do you get these clips bro

@Pirat_Nation WHAT THE FUCK

DDR5 and SSD prices at micro center Holy sh*t

@theo @SolutionB2u I laughed out loud when I read this reply lol "dumb ass with 2 followers" sounded so funny for some reason

@SolutionB2u I’m replying to your dumb ass with 2 followers, not much an echo chamber 🤷

Why have you closed your replies? It just creates an echo chamber where you only hear praise and shut out any real criticism.

@realprimolevi x.com/7a7zz/status/2…

@7a7zz Don’t you feel the models are heavily quantized?

got opencode go :)

@Anavarolik @ZDi____ its a pretty expensive model , use minimax and kimi k along side to save usage

@ZDi____ @7a7zz Legit but you reach the limit very fast :( (on opencode, didn't try on z.ai)

@ceasarxuu yah but its a 10$ sub and you can get much much more than 10$ s of usage so thats a win in my book :)

@7a7zz For high-intensity development, it's still a bit insufficient — the weekly quota is used up in about two days. Looking forward to a larger plan. Also, as I recall, opencode uses the full model without quantization.

@LearnWithScribe yah my guess is cuz glm 5.1 just came out everyone is trying to use it and overwhelming the inference

@7a7zz It was barely usable yesterday

wow GLM 5.1 is slow on opencode go to be fair , other models arnt this slow so I think this is just cuz of heavy traffic

not too sure about that , Im having a great expirience using opencode TUI and also the desktop, it is true it can get memory hungry and other tools like claude code can go even higher in terms of memory use. from the popular harnesses (cc , codex , opencode , cursor) opencode isnt the best cursor is but opencode is that bad its pretty good. but I have yet to try the niche harnesses like droid and factory and stuff

@7a7zz @KkCarson92338 I don't know if they used AI to build it or something, but opencode is genuinely the worst harness ever built, also after heavy usage it goes up to 14GB+ of RAM usage and nukes my PC

that is true but with copilot you could have some prompts that run for like 50k tones and some prompts that run for like 100 tokens, imo it depends on how you use it , for me the github copilot sub worked well and the opencode go sub is working well too. right now you can get the best value for 20$ by getting both a opencode go sub and a copilot sub and using both of them inside opencode. you can use opencode go and its open models as work horses and use copilot and its frontier models like gpt 5.4 for big tasks

@7a7zz @omni1896837 That's just because they can't subsidize nearly as much as Microsoft lol, Microsoft charges you per prompt, not per token or whatever, so I've literally had 1B tokens runs on Opus for 4 cents. Opencode will never be able to compete with large corporations

@mattpocockuk this is much smoother with better tooling tho :) we need more great developer tooling around our AI workflows like this

This is essentially the same as a homebrew workflow I've used for a while: 1. Create a /prototype-ui skill 2. Give it instructions to create multiple radically different designs 3. Give it a picker component to toggle between them Bingo

@satoru34176 I will touch you

@7a7zz Please Dont Bully me 😭

Not a Frontend Dev But I thought of this concept and just loveddd it ..

no not really , they work great tbh , it does feel a bit lower quality but only a bit. my guess is its quatized to FP4 but models can still work great at FP4 with good inference. for 10$ and with the amount of usage you get Im ok with a tiny quality hit ngl. I still can get a ton of work done with it, and if I need more I have a copilot sub

@theo not the suits 😭

Welcome to the FIRST EPISODE of TBPNN (Theo & Ben Podcast News Network) jkjk we're calling it Nerd Snipe. Available on Youtube, Spotify, and coming to all your favorite podcast networks soon™

look man I respect you and all but , you are SO DEAD WRONG you have no idea. this is the best horizen game of the whole series better than horizen 1. and a HUGE upgrade from FH 3 FH4 and FH5 AND JAPAN , like this is a car guy's dream game , and they did really well with the map and the whole vibe. I get where you are coming from but that is not what horizen is about its not NFS its no like racing drifting everywhere horizen is a whole vibe. You play NFS when you need a quick shot of adrenaline , you play a horizen game specially horizen 6 when you want to chill back and relax after a long day of work , horizen is the type of game that is you default comfort game for a car guy. thats the game we always go back to. so yah :) lv ur content and respect you but this is by far the worst take you have had

Forza Horizon 6 is a game that I wouldn't even bother to pirate. Nothing new added, the same driving mechanics, the same graphics, the same vibe as Forza Horizon 5. This Japanese "world map" could've been just a DLC. Forza Horizon 6 doesn't have any uniqueness.

@shahidcodes for me it was pretty good I do a lot of my work using it rn , ill post more about my experience as I use it

@7a7zz How’s the quality

@DvmsoBBM @KkCarson92338 idk but Im pretty sure opencode is better than copilot as a harness

@7a7zz @KkCarson92338 Actually opencode consistently worsens the output quality on various benchmarks for both GPT and Claude models, I thought you guys already knew this

@DvmsoBBM @omni1896837 why is that , can you drop me a DM saying how exactly that happens and I will test and reach out to the openc9de team

@7a7zz @omni1896837 My usage dies after a few prompts, it's a pretty bad deal really, especially for inferior models

@prashant_yi kinda but its mostly cuz everyones using it

@7a7zz How's the glm 5.1 t/s ?

@ZDi____ pretty legit

@7a7zz How's GLM 5.1? Benchmaxxed or legit?