EmpirioLabs AI model provider for Hermes Agent

EmpirioLabs AI is a specialized AI inference and integration provider. It hosts open-source models on GPU infrastructure it manages, runs optimized endpoints for proprietary models, helps teams ship their own models, and offers on-demand GPU Cloud instances and hosted AI agents. Its model catalog spans text, image, audio, video, transcription, embedding, reranking, research, and 3D models.

This plugin adds EmpirioLabs AI to Hermes Agent as a model provider named empiriolabs. For image generation, video generation and text-to-speech, install the companion plugin empiriolabs-media.

Requirements

Hermes Agent 0.21.6 or later
An EmpirioLabs AI API key from https://platform.empiriolabs.ai/dashboard/api-keys

Install

From the Hermes plugin catalog:

hermes plugins install empiriolabs

From GitHub (add --ref <full commit SHA> to pin a reviewed commit):

hermes plugins install EmpirioLabs-AI/hermes-empiriolabs/provider

The installer asks for EMPIRIOLABS_API_KEY and saves it to your Hermes .env file. You can also add it there yourself.

Use

Run hermes model, choose EmpirioLabs AI and pick a model from the live catalog. In a session, /model switches models the same way.

Run a one-shot prompt:

hermes -z "Summarize the README in this folder" --provider empiriolabs -m glm-5-3

Or set it in config.yaml:

model:
  provider: empiriolabs
  default: glm-5-3

What the plugin provides

Model list. Chat models that support tool calling, read from GET https://api.empiriolabs.ai/v1/models. Hermes sends tools with every request, so models without tool calling are left out. When the catalog cannot be reached, Hermes offers two fallback ids (glm-5-3, qwen3-7-flash).
Model details. Context window, image input and reasoning support for each model, from the same catalog.
Auxiliary model. Hermes uses a low-cost model for side tasks such as context compression and titles: qwen3-7-flash, the lowest-priced featured chat model with tool calling and at least a 200K context when this version was released. Where Hermes asks the provider for a current pick, the plugin applies the same rule to the live catalog.
Reasoning effort. When you choose a level with /reasoning, the plugin sends reasoning_effort to models that accept it, using the closest level the model offers without going higher. Nothing is sent when no level is set.
Endpoint override. Set EMPIRIOLABS_BASE_URL to use a different API base URL (default https://api.empiriolabs.ai/v1).

Data handling

Hermes sends your prompts, messages, tool definitions and tool results to api.empiriolabs.ai (or EMPIRIOLABS_BASE_URL) with your EMPIRIOLABS_API_KEY. Usage is billed to the EmpirioLabs AI account that owns the key.
The plugin itself only reads the model catalog (GET /v1/models) and keeps it in memory for an hour. It writes no files, starts no background processes, runs no shell commands and sends no telemetry. It never reads or stores your key; Hermes does.
EmpirioLabs AI privacy policy: https://empiriolabs.ai/privacy-policy

License

MIT

empiriolabs

What it adds

Environment variables it needs 1

README