openclaw

has anny one herd of openclaw.
if so, this is a discusion about openclaw and your expirences

hermes is better

yeh, at the time, i did not kno about hermes

Hermes is decent, but there are some cases where openclaw is better. For instance, when you do cron in Hermes, you can't have crons fire into the session you're working in, each cron gets its own session and there's nothing you can do about it. Openclaw allows you to have cron jobs fire into the main session, which is really nice if that job requires context from said main session or you want to just refer back to it and you want your agent to have the context without it checking the cron output first.

Honestly, I just use Chatgpt pluss because for what it is worth it is ausom. But I also have Hermes and it is pretty desent too.

chatgpt pluss from cookie

Can somebody explain why it would be useful to have, say, Hermes or Openclaw alongside my AI subs? Most of the time I feel like I'm using AI too much anyways, but I don't see a reason to have Hermes running on my machine.

If you want more than just coding agents and you want it to do more than that and have more permanent context on you to help you with stuff, Hermes or Openclaw is the way to go. Codex and claude code are more coding focused, Hermes and Openclaw are the I want to use my AI for more than just coding options.

I use opencode free, I am considering using Hermes free as well to increase my free options, is that viable?

It's possible to use Hermes for free, but the rate limits aren't going to be fun. You won't be getting long coding sessions out of it that's for sure.

i've had no rate limit problems

I guess it depends on what model you use, and also what time you use it. There are peak hours and some of my US friends have run into rate limits almost straight away.

What are your best models on there?

i use the built in models for hermes

What are the limits like? I mean for real blind use personal experience

Replying to Natan44That does not contribute to the topic. And also no it is not from him.

i no it's from him, alfie told me

@Destructatron: Okay, you can do more stuff. But what exactly? For instance, why would Hermes me beneficial for me if I use Gpt/Claude for chat and Codex/Claude Code for work/coding? What actual benefit does it provide if I run Hermes on my local machine? I'm inclined to try it but I guess it takes a while to get used to and setup everything, and I'm not willing to pay API rates so I'd have to integrate my sub, which just means less limits overall.

@17 No its not. I can show you if you want.

The immediate benefit for me is that I don't have to type crap into a terminal and deal with terminal stupidity when I want to paste something, read messages, look back at history, etc. You can connect Hermes to Discord, Telegram, Signal, WhatsApp, basically any channel you want. Claude code and codex can do this but it requires more setting up and there's no first class support for it. Also, Hermes updates its memory and creates new skills over time, so you don't have to keep telling it to do these things, and it tells you when these things happen as well. It now also has the ability to connect natively to a lot of apps you use daily like gmail and other google stuff, with the potential to add way more. If you don't want to deal with the stupid gmail interface for example, go ask your agent to retrieve your emails and tell you if there's anything worth your attention, and you can make this a cron job too, basically a scheduled task that repeats at certain times. I'm only literally scratching the surface of what agents in general can do here. Also, if you use Chat GPT Hermes can sign into that by itself, and Open AI has said publicly that they allow the use of the auth token in agent frameworks like Hermes, Openclaw, etc.
-- (Urh2006):
@Destructatron: Okay, you can do more stuff. But what exactly? For instance, why would Hermes me beneficial for me if I use Gpt/Claude for chat and Codex/Claude Code for work/coding? What actual benefit does it provide if I run Hermes on my local machine? I'm inclined to try it but I guess it takes a while to get used to and setup everything, and I'm not willing to pay API rates so I'd have to integrate my sub, which just means less limits overall.
--