Thank you Martijn Smit for being my first annual customer for LinkBunny.
Black Friday deal is up. Get 30% off your first annual payment with BLACKFRIDAY2026 ($147, then $210/year). First-time paying customers only.
https://t.co/i0cDc5Ugvd
Strange how multiple X ad reps email me to want to talk, but none reply to me when I ask them questions about campaigns I'm running. It's a small account for sure, but you emailed me first 🤣
The best deal is now Claude Code with Opus 5.5. And I think generally best to spend on $200 subs (or $100 if lower budget) rather than spread it out, unless there are really more specific reasons. If even lower, probably something like DeepSeek ($200 subs are cheaper)
I don't know what to do with Codex.
ChatGPT is my favourite speak to AI by far.
I use it to code too (not much latelly because the lack of time).
I use it to review code.
But still not spend it all.
You know, what if they are the fairest about price?
I don't feel like I should downgrade it.
Any Ideas about what Astra is great at?
Videos, Images, Voices?
@JesusAnutrof If you give them similar tasks, you can tell Codex to compare them by studying the sessions for effectiveness, efficiency and tps after a few days.
@JesusAnutrof > But it spends faster than codex.
It's output tps is 3x unless you use Ultrafast. So while per token-wise it is slightly more expensive than Sol, it is noticeably more expensive only because it finish more work, faster.
@GergelyOrosz@sean_j_roberts Sorry to hear about your horrible experience, but this is an unintended pun about Atlassian's "DX" right? —
> not about Atlassian really but DX