@ydennisy_ you can use your codex sub with pi, though! i pair a codex sub with hf inference providers and it's REALLY good.
you can use open models that are 20x cheaper than closed models on pay-as-you-go api. and for heavy lifting, you just switch to gpt-5.5.
works like a charm
@oknwook i just saw that :( to be fair, it was ridiculously charitable. the usage of the previous plans was even measured in requests rather than tokens π
You can run Hermes Agent on your own VPS for about $15/month.
I recorded a quick walkthrough of the full setup: cheap VPS, Minimax token plan, and an AI agent skill that configures the server for you.
@testy_cool great to hear mate π€ more coming up. notably, take a look at this setup right here: https://t.co/H2GsVlcTow with a single command you can run hermes on a hf bucket. even more affordable than this!!
@NoeFlandre oh dang... i must have recorded this just a few minutes before they changed it π such a pitty, i think this was the absolute best plan for running a claw on a budget :S
the recording of my talk at prague crawl on creating LLM datasets is out!
it covers a high-level overview of the process to create the fineweb dataset in about 20 minutes π€
you can watch it here π