A heads-up display (HUD) with per-turn telemetry for both local inference engines and remote models. Look for it in the sidebar.
opencode tui hud mlx inference-engine ttft llama-cpp vllm koboldcpp sglang opencode-plugin mtplx opencode2 tokens-per-secon mlx-serve
-
Updated
Sep 23, 2026 - TypeScript