One team, your providers
Keys and subscriptions belong to your team. Pick a default model and harness, then override them per project or chat.
Features
Runs bill to your own account first, on the harness each team, project, or chat picks.
Built for an open stack
Connect a key or subscription once, and matching model calls use it before platform FTAI.
Keys and subscriptions belong to your team. Pick a default model and harness, then override them per project or chat.
Add OpenAI, xAI, or OpenRouter API keys. Each is encrypted and stored for your team.
Use a ChatGPT Plus or Pro, or
Run turns on the built-in runtime,
If your key is unavailable, calls fall back to platform FTAI only when you turn that on.
Works with your chosen stack
OpenAI, xAI, and OpenRouter, by API key or account sign-in.
Keys are encrypted at rest, scoped to your team, and every read is logged.
Set the default harness for a team or project, and override it for one chat.
Tools that change data in a connected account wait for a person to approve.
Pick a model per chat, or set team defaults for chats and subagents.
See spend grouped by model, including which model each call was routed to.
Which providers and harnesses you can use, and where your keys live.
Want to review your model and harness setup? Talk to us
Connect your keys or subscription, pick a harness, and run on your own account.