Los Angeles, CA
Joined September 2020
Hey @Square we need a Grok Bot MCP
13
Bro GPT 6.1 sol has been so slow the past 3 hours wtf
1
6
5,982
Should I go for 64GB or 128GB of RAM on a MacBook Pro M5 Max? 🤔 I'm upgrading from my 16GB M1 MacBook Pro, and I'm seriously torn between the two. I run multiple AI coding agents, development tools, and other heavy workloads simultaneously. Is 128GB worth the extra cost, or is 64GB more than enough? Would love to hear from anyone with a similar workflow.
192
5X plan running Sol 6.1 medium. Reset only lasted a day🫩
1
44
Ran out of codex usage and here is grok to save the day.
1
1
106
I have 4 projects that will be releasing soon👀
10
Claude Code’s /advisor is kinda wild: I’m running Opus 5.5 as the main model with Fable 5.1 on call. Opus does the work, but when it hits an ambiguous/critical decision or gets stuck, it can escalate server-side to Fable for guidance, then keep going. Lowkey this is amazing! 🤯
38
Current $200 plans: Claude 20x • Opus 5.5 • Heavy usage • Still lasts noticeably longer Codex $200 • Astra • Heavy usage can burn through the allocation in 1–3 days Same $200 price point. Very different usable capacity. OpenAI needs to make the Codex plan more competitive.
1
86
Andres Eger retweeted
Introducing Vitals for Mac  The Activity Monitor alternative that thinks in apps, not processes. $5 for a lifetime license. 38 left. > vitalsmac.com
128
38
36
519
146,859
Opus 5.5 is indeed the best I have seen
28
Insane that i have been using Opus 5.5 since it came out and its barely consuming usage 👀
28
Claude giving a banked reset
32
I’ve been using Gemini 3.8 for the past couple of days, and honestly, I’ve gotten really good results with the kind of work im using it for. ( Specially with /boost )
33
Grok seems very slow today wtf
20
Picking up 2 DGX Sparks tomorrow. Time to make GLM 5.3 Flash and DeepSeek my daily drivers.
3
55
Claude maxed out, Codex usage gone too 😭 Grok it is for now.
1
1
49
Current status: locked in🔒
1
20
So Uh... is @AnthropicAI going to boost it again?😅
22
The new @ollama support in Codex is so good. Hit my usage limit, switched to GLM 5.3 Flash in seconds, and kept working. Chef’s kiss 🤌
79
The biggest mistake with GPT-6 Astra in Codex is overprompting it. • Fewer instructions • Shorter skills • Leaner AGENTS.md • Clear decision boundaries for tasks • Clear definitions Astra High has also used the least usage in my testing so far!
1
1
50