Parallel and cloud environments
Harbor abstracts the environment behind a provider. The same task moves from your
laptop to the cloud by changing --env.
# local Docker (default)harbor run -d terminal-bench@2.0 -a claude-code -m anthropic/claude-opus-5-5 -n 4
# Daytonaharbor run -d terminal-bench@2.0 -a claude-code -m anthropic/claude-opus-5-5 -n 100 -e daytonaProviders listed upstream
Section titled “Providers listed upstream”Daytona, Modal, LangSmith, Blaxel, Novita Sandbox, Tensorlake, Runta, Sail.
Why parallelism matters
Section titled “Why parallelism matters”- One task exposes one failure mode; a thousand tasks expose a distribution
- RL needs rollouts, and rollouts need throughput
- Re-running across providers is how you test whether the task itself is sound