Some models like Gemma are good and they seem to be worth using for coding and the reasoning chats. It's worth trying them over openrouter.
Some models like Gemma are good and they seem to be worth using for coding and the reasoning chats. It's worth trying them over openrouter.
I have been using gtp-5.4-mini with Hermes as lately after got tired of Openrouter latency problems, some days its great some others not much so I had to swap for gpt oatuh, been running great only that it takes from my Codex sessions, doesnt take that much
You can also try the ollama on hermes. I think GLM 5.1 is a good model, or minimax m2.7 or even Kimi K2.6