Built one of the top voted Discord bots on top.gg (80k+ servers). Now building Sirat, GuildMate & more
United Kingdom
- Tweets 471
- Following 262
- Followers 29
- Likes 83
6.1 Sol is really efficient
Gave it a task and ran it on high
37 minutes, used only ~16% of my 5hr usage
Decently usable on a $20 plan, esp if you use low/med
2
11
Ok we don’t even get it anyway right now
I hope they give us the same 1 month no usage consumption when it does roll out @OpenAI 🥲
2
19
I’ve been struggling to build macOS apps on my 8GB M1 MacBook. Have to close everything else and even then I run out of memory
Finally started using @cognition Devin's cloud Mac. I get an M4 Pro and 16GB RAM!
Even gives me an attachment for the build!
3
46
I’ve been trying @paseo_sh and I think it’s gonna replace @orca_build for me
I used to hate chat UIs (used only terminal) but I’m starting to change my mind
Even when I still want to use terminal, Paseo lets me do that
It’s missing a couple things but oss so I can tweak it
1
3
22
If you’re not using SimCord for Discord bot dev (discord py), you’re really missing out
3.0 is gonna be huge too
1
2
33
I wish Grok 4.7 came with the higher context window tier as default in Cursor
And if you have to charge extra, just do it for tokens over 256k
1
66
Hey @cursor_ai I mistakenly upgraded plans, haven’t used any of it except for Grok Bot jobs that ran automatically (1% weekly used)
Customer support AI is saying I’m ineligible for a refund which I think is cos of the @bot thing? Would appreciate some help 🥲 @ericzakariasson
3
46
Ran this through GuildBench (1 pass), my benchmark that tests Discord operations. Results are actually great, especially at this price
Mercury 2.5 was the speed/cost winner: 4.4s median and ~$0.0005 per successful journey
Honestly surprised a diffusion model can do this!
1/2
Made with AI
1
2
135
I’m starting to think ADR docs are an anti pattern for new projects. The architecture and design is often rapidly changing at the start and it locks the AI into one path. I have to keep telling the AI to challenge our conventions if they don’t make sense
1
2
48
Did anyone rate GLM 5.3 this high? It seems like the Flash variant is surprisingly good but shouldn’t 5.3 perform better than Flash?
The new AI model that’s swept to the top of online usage charts with high performance at zero cost was created by China’s Z.AI, also known as Zhipu bloomberg.com/news/articles/…
5
127
I know people are complaining about the 5hr limits coming back but I think I like it? Helps me pace my usage
2
28
Trying out explicit tags when I don’t want the AI to over-eagerly implement
What do you think about x [QUESTION]
Could we try y [ACTION]
2
22
Practically unlimited usage
Need to try this for myself. Someone get this benchmarked too!
Ox Alpha (stealth model) is free for the next week
- 1M Context
- Multi-modal
- Zero Data Retention
Generous rate limits, near unlimited usage
We have capacity for 100T tokens per day, lets see what you can do
2
127