What You Get
TypeScript & Python
Official clients for the two languages that matter most in AI. Full type safety, async support, and idiomatic APIs for each ecosystem.
Drop-In LLM Metering
Swap one import and every LLM call is automatically metered and attributed. Works with OpenAI, Anthropic, and any OpenRouter-compatible provider.
Wallet Snapshots
Query wallet balances from your application code. Let agents check their own budget before making expensive calls — enabling cost-aware reasoning.
Per-Request Settlement
Each API call settles atomically against the user's wallet via x402. No batched invoices, no webhook reconciliation, no billing edge cases.
Session Key Management
Ephemeral signing keys that auto-zero on disposal. Defense-in-depth key handling so your users' funds stay secure even if a session leaks.
Streaming Support
Full support for streamed LLM responses. Token usage and cost are tracked accurately even for long-running streaming completions.
