Start free in Observer Mode to see where workflows decay. Move to Pilot when you're ready for active governance.
See where workflows decay before you buy.
Note: A workflow run is one full orchestrator execution monitored by Luthy. One run may include many LLM calls.
For teams ready to turn on Luthy in real workflows.
Overage: $8 per 1,000 additional LLM calls
For production agent teams with sustained usage.
For scaled deployments and platform teams.
Luthy bills on LLM calls evaluated.
A single workflow run may contain many LLM calls. That's why billing is based on evaluated calls, while workflow runs are shown only for reporting during the observer trial.
All evaluated calls count toward usage, regardless of environment.
No. Luthy operates as an external runtime layer that observes execution behavior without altering model weights or requiring retraining.
Wrap your existing agent code with Luthy's protection layer. No infrastructure changes or model modifications required.
Luthy works with Python-based agent systems and any model API that your stack already calls.
Start free in Observer Mode and see where Luthy would detect non-progress.