AxonSDK

AxonSDK provides a universal interface for routing AI inference requests across multiple compute providers, including GPU clusters, TEE nodes, edge networks, and major cloud platforms. It is designed for developers who need to deploy AI inference workloads flexibly without altering their application logic when switching between different providers or environments. The SDK is available in both TypeScript and Python, maintaining consistent configuration and method patterns across both languages to simplify cross-language development.

Key features of AxonSDK include provider-agnostic routing, allowing users to direct inference to backends such as io.net, Akash, Acurast, Fluence, Koii, AWS, GCP, Azure, Cloudflare, and Fly.io through a single integration. The SDK supports smart routing capabilities such as circuit breaking, health monitoring, automatic failover, and live provider pricing signals to optimize routing decisions. Security measures are built in, with protections against SSRF, enforced HTTPS-only egress, and DNS-rebinding defense on every request.

AxonSDK offers a drop-in OpenAI-compatible inference handler, making it possible to redirect existing OpenAI client calls to its unified endpoint without code changes. The TypeScript SDK is compatible with Node.js, Next.js, Cloudflare Workers, and React Native, while the Python SDK is async-native, type-checked end to end, and compatible with Python 3.11 or higher. Both SDKs allow developers to move between languages without rethinking their inference layer.

The project is open source under the Apache-2.0 license, free to use with no telemetry or vendor lock-in. Users have the option to fork, audit, or self-host the router as needed. AxonSDK is intended for teams and developers seeking a flexible, provider-neutral solution for deploying AI inference workloads across diverse environments.