How to Configure Local Models for Hermes Agent

Dry_Long3157 · reddit · 2026-07-14

Running a local setup with a 4090 + 5060 Ti and 64GB of RAM, the author is using opencode with MiMoV2 Pro but finds the $20/day cost too high.

They are asking if anyone has successfully run an open-source model on this hardware via Hermes Agent to actually complete tasks at reasonable speeds. They also mentioned hearing that Qwen3.6-27B causes looping issues in Hermes and are looking for viable alternatives.

Original post →

More from coding & agent

coding & agent channel →