OpenNodes

io.github.opennodes-iov0.1.2Updated Oct 4, 2026

Discover verified AI inference nodes, get the best node+model per task, pre-price chains.

VerifiedSTDIODesktop onlyDeveloper ToolsAI & ML

Overview

AI-generated overview

Lets an assistant discover verified AI inference nodes, pick the best node and model for a task, and pre-price multi-step chains.

What it does
A local stdio MCP server from the OpenNodes Protocol client. Its tools include recommend, which returns the best node and model for a task with pricing and an explanation, plus search_offerings, get_offering and estimate for browsing and pre-pricing offerings. The same package also ships a CLI and a local OpenAI-compatible gateway that routes requests across discovered nodes.
When to use it
Useful when you want an assistant to choose among many AI inference providers or models by task, compare prices before running a chain, or route OpenAI-SDK style requests through a local gateway. Less relevant if you already use a single fixed model provider.
Requirements
Runs locally via npx from the npm package @opennodes/cli, so Node.js is needed. It talks to a registry; the ONP_REGISTRY environment variable sets the registry base URL and defaults to the hosted registry, or can point at a local registry. No authentication is declared.
Before you install
The recommend tool sends extracted features of your prompt to the registry rather than the prompt itself, so review what is extracted before using sensitive input. The gateway and invoke paths spend money on inference and produce signed receipts. The registry endpoint is configurable through ONP_REGISTRY, so confirm it points where you intend.

Installation

In SourceWeft

  1. Open OpenNodes in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

@opennodes/cli

The OpenNodes Protocol (ONP) client: discover verified AI inference nodes, pre-price multi-step chains with enforceable estimates, invoke with pinned prices and signed receipts — and bring all of it into the tools you already use.

  • onp mcp — a stdio MCP server for Cursor, VS Code, Claude Desktop and any MCP host. Tools: recommend (best node + model for a task, priced and explained; the prompt never leaves your machine — only extracted features are sent), search_offerings, get_offering, estimate.
  • onp gateway — a local OpenAI-compatible server over discovered nodes. Point any OpenAI-SDK app at it. Virtual models onp/auto, onp/auto-cheap, onp/auto-fast, onp/auto-quality, onp/auto-private route each request by the advisor; aliases and query-defined virtual models are supported; every call is receipted.
  • onp search | estimate | invoke | card — the CLI.
bash
npx @opennodes/cli mcp --print-config      # paste into your MCP host's confignpx @opennodes/cli gateway --port 4141     # then set OPENAI_BASE_URL=http://127.0.0.1:4141/v1npx @opennodes/cli search --modality text --tier verified --sort tps

Registry: ONP_REGISTRY — defaults to the hosted registry at https://registry.opennodes.io (9,000+ offerings); set it to http://127.0.0.1:4300 to use your own (npx @opennodes/registry).

Part of OpenNodes · specification · Apache-2.0.

Source: impl/packages/cli/README.md at commit fea7b17

Tools

0
Tool metadata has not been indexed yet.

Version history

1
  1. v0.1.2LatestOct 4, 2026