The product is an API and platform for running AI sessions that routes transcripts, model selection, and compute through one interface and allows switching models or machines mid-session without losing context. It is for software developers and engineering teams building AI agents and app backends, with use by finance and security teams in startups to enterprises managing AI spend and audit trails. It is delivered as an API, CLI, and dashboard under a bring-your-own-keys model with vault integrations, routing policies, and per-session cost attribution.
Key features
- One API for transcript, model, and compute
- Switch model and compute mid-session
- Create, run, peek, terminate sessions
- List, attach, switch computes
- Project namespaces for keys and billing
- Vault integration for Infisical and Doppler
- Routing policies for model choice
- Headless chat embed without UI work
- Multi-tenant per-user scoped sessions
- Cross-model widget for any model
- CLI commands run, chat, compare
- Dashboard for sessions, usage, billing
- Per-run cost attribution and caps
- Transcript audit trail for every run
- Client-safe publishable keys for browser
- Local or cloud compute attachment
- No social media activity within the last 30 days
GTM channels
- Blog
- API
- Docs
ICP
- Software developers
- Engineering teams
- Startups