Privacy policy

1. What we collect

• Sign-in details. When you sign in with GitHub or Google we receive your verified email and your provider account ID from our identity provider, Clerk.
• Workspace data. Workspace membership, API key hashes, the Wallet ledger, Jobs, Trials and their Rewards and costs.
• Your content. Tasks and files you upload, private Benchmarks, guided-authoring conversations, and Trial transcripts.
• Technical data. Server logs, IP addresses (used for rate limiting and abuse prevention) and basic error reports.

2. How we use it

We use this data to operate the service, bill your Wallet, prevent abuse, keep the service secure and fix faults. Uploaded files, private Benchmarks and transcripts are used only to serve your Workspace.

3. Who else sees it

• Model Providers. To run a Trial, the Instruction, Environment output and related text are sent to the Provider that serves the Model, either directly (Anthropic, OpenAI, Google) or via OpenRouter. For private Benchmarks, Workspaces can restrict which Providers are allowed, and OpenRouter requests ask not to collect data.
• Infrastructure. Sandboxes run at Daytona. Sign-in is handled by Clerk. Hosting and backups run on our own servers and a backup storage provider.
• Staff. Staff cannot see private Benchmarks unless your Workspace has turned on staff access for done-for-you work, and those actions are audit-logged.

We do not sell your data.

4. How long we keep it

• Trial transcripts are deleted after 90 days.
• Rewards, costs and aggregates are kept, as they are the record of what you were charged.
• You can delete a Job’s transcripts early from your Workspace.
• Expired Quotes, device codes and unused upload archives are deleted automatically.

5. Your choices

You can ask to see, correct or delete data we hold about you, subject to what we must keep for billing records. Contact us using the link below.

6. Contact

Open an issue. A dedicated privacy contact will be added before launch.