Per-client configuration

AI models

Choose which model runs each Wave AI task: Tidal Wave's Claude, Claude on your own Anthropic key, or your own OpenAI-compatible endpoint.

Where your data goes

When a task runs on a model you choose, Wave sends that task's instructions and your workspace's data to that model's provider, under your own agreement with that provider. Tasks you leave on Tidal Wave's Claude keep running there.

Open-web research and embeddings always stay on Tidal Wave's platform.

Select a workspace to manage its AI models.