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 repository holds two plugins for Hermes Agent:
| Plugin | Directory | What it adds |
|---|---|---|
empiriolabs |
provider/ |
A model provider: chat models with tool calling from the live EmpirioLabs AI catalog |
empiriolabs-media |
media/ |
Backends for Hermes' image generation and editing, video generation and text-to-speech tools |
Both read their model lists from GET https://api.empiriolabs.ai/v1/models at runtime, so new models appear without a plugin update.
- Hermes Agent 0.21.6 or later
- An EmpirioLabs AI API key from https://platform.empiriolabs.ai/dashboard/api-keys
From the Hermes plugin catalog:
hermes plugins install empiriolabs
hermes plugins install empiriolabs-mediaUntil the catalog entries are live, install from GitHub (add --ref <full commit SHA> to pin a reviewed commit):
hermes plugins install EmpirioLabs-AI/hermes-empiriolabs/provider
hermes plugins install EmpirioLabs-AI/hermes-empiriolabs/mediaThe installer asks for EMPIRIOLABS_API_KEY and saves it to your Hermes .env file. Enable the media plugin when asked, or later with hermes plugins enable empiriolabs-media.
Chat: run hermes model, choose EmpirioLabs AI and pick a model, or pass the provider on the command line:
hermes -z "Hello" --provider empiriolabs -m glm-5-3Image, video and speech: run hermes tools, open Image Generation, Video Generation or Text-to-Speech, and select EmpirioLabs AI. The same settings in config.yaml:
model:
provider: empiriolabs
default: glm-5-3
image_gen:
provider: empiriolabs
model: flux-2-klein-4b
video_gen:
provider: empiriolabs
model: hunyuan-video-1-5
tts:
provider: empiriolabs
model: tts-2-flash
voice: SarahEMPIRIOLABS_BASE_URL overrides the API base URL (default https://api.empiriolabs.ai/v1). Each plugin's README covers its options: provider, media.
- Prompts, messages, tool definitions and results, source images and text to speak are sent to
api.empiriolabs.ai(orEMPIRIOLABS_BASE_URL) with yourEMPIRIOLABS_API_KEY. Usage is billed to the EmpirioLabs AI account that owns the key. Image, video and speech calls are billed when the agent makes them, without a separate Hermes approval prompt. - Generated media is downloaded from the URLs the API returns into Hermes' cache directory. The key is never sent to those URLs.
- The plugins read the key through Hermes and never store it. They write no files other than generated media, read no files other than a local image you ask Hermes to edit, start no background processes, run no shell commands and send no telemetry.
- EmpirioLabs AI privacy policy: https://empiriolabs.ai/privacy-policy
Run the tests in an environment where Hermes Agent is installed (or set HERMES_AGENT_ROOT to a checkout):
python -m pytest
hermes plugins validate provider
hermes plugins validate mediaAll HTTP in the tests is mocked. scripts/live_smoke.py makes real, billable requests (about one cent): it lists models, sends one short chat, generates one image and, with --tts, synthesizes one sentence, then prints what the run cost.
EMPIRIOLABS_API_KEY=... python scripts/live_smoke.py --tts- Website: https://empiriolabs.ai
- Documentation: https://docs.empiriolabs.ai
- API keys: https://platform.empiriolabs.ai/dashboard/api-keys
- Support: [email protected]