Comment by simonw

Anyone seen a low-friction way to run prompts through this yet, either via a hosted API or chat UI or a convenient GGML or MLX build that runs in Ollama or llama.cpp or LM Studio?