📊 Leads every instruction-following benchmark and nearly all tool-use benchmarks against models up to 4x its size, trailing only #Qwen3_5 9B on BFCLv4 — coding remains the weaker area🔧 Day-one support for #llamacpp (GGUF),
#OpenCode with #Llamacpp — How it Works in Practicehttps://medium.com/rigel-co…
OpenCode with Llama.cpp — How it Works in PracticeFour components, one chain — and a proxy that holds it all together
@michabbb@social.vivaldi.net
@frankel@mastodon.top