hax

hax is a terminal-native coding agent shipped as a single native C binary that starts instantly and uses only a few megabytes of memory, leaving more memory for local models. It auto-discovers the model and runtime capabilities of a local setup, so a llama.cpp server works without a custom provider configuration block, and it also reaches hosted providers such as OpenAI, Anthropic, Codex through a ChatGPT subscription, OpenRouter and OpenCode Zen. The interface streams markdown and live tool output while preserving native scrollback, and a transcript view plus an optional wire-protocol trace let developers inspect exactly what was sent to the model and what came back.