Documentation
AgentMesh docs
AgentMesh runs open-weight AI models on computers contributed by independent providers, behind an OpenAI-compatible API. These guides cover using it, hosting it, building on it, and running your own network.
Getting started
What AgentMesh is, who it is for, accounts and roles, and the live links.
Customer guide
Deploy a model, call it with an API key, and understand billing.
Provider guide
Install the worker, enroll machines, set their prices, manage them, and get paid.
API reference
Authentication, models, deployments, inference, streaming, keys, errors and usage.
Operators & self-hosting
Run your own AgentMesh: configuration, deployment, catalog, pricing, payments and recovery.
AI deployment agent
The Gemini-powered deployment agent: how it works, configuration, limits and behaviour when Gemini is down.
Security & limitations
Trust model, data visibility, controls, residual risks and what is not verified.
Quick start
- 1Deploy a modelSign up as a customer, pick a model, deploy, and call it with an API key.
- 2Host modelsSign up as a provider, create an enrollment token, and install the worker.
- 3IntegrateUse any OpenAI client against your deployment’s /v1 endpoint.
Live links
- Dashboard
- www.agentmeshx.tech
- Become a provider
- www.agentmeshx.tech/register?role=provider
- API base URL
- agentmesh-riuz.onrender.com/api/v1
- Worker releases
- github.com/meshach5667/AgentMesh/releases
- Source code
- github.com/meshach5667/AgentMesh
Your requests run on a computer owned by an independent provider