Adapters for deploying Infernet provider nodes to cloud GPU services (RunPod, vast.ai, Lambda, etc.).

INFERNET SDK - Agent Registry for Autonomous Intelligence

JavaScript/TypeScript SDK for the Infernet Protocol REST + SSE API.

INFERNET CLI - Agent Registry for Autonomous Intelligence

Nostr-based authentication helpers for the Infernet Protocol.

Inference engine abstraction for Infernet — pluggable backends (Mojo+MAX, in-process stub) over a stable async-iterator API.

Shared constants for the Infernet Protocol — supported payment coins and platform deposit addresses.

OpenAPI 3.1 specification for the Infernet Protocol control-plane API.

GPU-node CLI / daemon for the Infernet Protocol — register a GPU server with a Supabase control plane and start earning.

CoinPayPortal payment gateway integration for the Infernet Protocol.

Structured logger used across Infernet Protocol packages.

Distributed inference coordinator and worker for the Infernet Protocol.

NVIDIA NIM / build.nvidia.com adapter for the Infernet Protocol — treats NIM's OpenAI-compatible endpoint as a fallback inference provider.

Supabase-backed model layer for the Infernet Protocol (Provider, Client, Job, Aggregator).

Local hardware detection for the Infernet Protocol — GPUs (NVIDIA, AMD, Apple Silicon) via native tooling, plus CPUs / RAM / arch via node:os. Pure-JS for the CPU side, vendor-CLI shell-out for GPUs.

Training orchestration for Infernet — pluggable backends (DeepSpeed/FSDP for trusted clusters, OpenDiLoCo for cross-provider async, Petals for trustless P2P fine-tunes) over a stable async-iterator API.