callmej
451 posts


@World_Affairs11 No they are not, this is past news. Dont make it seem like its happening now.
English

the longer your Hermes agent session runs... the more every message costs.. I dont know if you know this but there's a command that fixes this but most people don't really know it even exists
/compress
It summarizes your entire conversation history, keeps the important context, and slashes your token count with getting lost on what it was working on
share it with a buddy that burns tokens like no tomorrow
English

@bidcapbro Drone attack against a NATO member state like Romania does trigger Article 5.
English

@hackyguru @mart1n I use ollama, the $20 plan and even the $100 plan go a long way. Its so cheap. DeepSeek is pretty dope too.
English

@bidcapbro @mart1n do you use the hermes portal or glm inference directly in a cost pov?
English

GLM-5.1 is quietly becoming one of the best open source models for agentic workflows. Long-horizon tasks, multi-step debugging, extended sessions.
It handles what many GPT/Hermes setups fail at. Worth trying. #AI #OpenSource
English

@mart1n tried supergrok 1M context with hermes. nowhere close to sonnet/gemini/opus :/
English













