Getting Started

OpenSyntax starts with provider onboarding when no provider is configured. After setup, you can run it from any project folder.

First launch

opensyntax

If no provider exists, OpenSyntax shows a terminal setup wizard with provider choices such as OpenAI, Claude, Gemini, OpenRouter, NVIDIA NIM, Ollama, and LM Studio.

Connect a provider

opensyntax auth

Select a provider, choose API key authentication when prompted, enter the key securely, select a default model, and validate the connection.

Use an API key directly

opensyntax config --provider openai --model gpt-4o-mini --api-key "$OPENAI_API_KEY"

Validate setup

opensyntax doctor

The doctor command checks provider reachability, model discovery, chat streaming, git availability, and workspace state.

Start in a project

cd path/to/project
opensyntax "explain this repository"