Hermes Desktop
About
Hermes Desktop is a native desktop app for installing, configuring, and chatting with Hermes Agent. It provides a GUI for provider setup, model selection, sessions, tools, skills, and messaging gateways — without having to manage the CLI by hand.
As of PR #612 (merged June 9, 2026), AI/ML API is a built-in provider preset in Hermes Desktop. You can select it from the provider list during first-run setup or at any time in Settings.
Quick start
Download and install Hermes Desktop from hermesagents.cc.
Open the app. On first launch it runs the setup wizard.
When prompted for a provider, choose AI/ML API.
Enter your AI/ML API key.
Pick a model and start chatting.
When to use AI/ML API with Hermes Desktop
AI/ML API works well with Hermes Desktop when you want:
one key for many model providers
an OpenAI-compatible endpoint with no custom configuration
fast switching between coding, reasoning, and multimodal models from a GUI
Prerequisites
Before you start, make sure you have:
Hermes Desktop installed (hermesagents.cc)
an AI/ML API key from aimlapi.com/app/keys
a model ID from aimlapi.com/models
The base URL is:
Need a key first? Use API Key Management.
Set up AI/ML API in Hermes Desktop
First-run setup wizard
When Hermes Desktop opens for the first time:
Choose Local mode (runs Hermes Agent at
127.0.0.1:8642) or Remote mode (connects to a hosted Hermes API server).At the provider step, select AI/ML API from the list.
Enter your
AIMLAPI_API_KEYin the key field.Pick a model from the auto-discovered model list.
Complete setup. The app saves your config and opens the workspace.
Change provider in Settings
To switch to AI/ML API after initial setup:
Open Settings → Provider.
Select AI/ML API.
Enter your
AIMLAPI_API_KEY.Save.
Saved configuration
Hermes Desktop stores provider config in the standard Hermes files under ~/.hermes/:
~/.hermes/.env:
~/.hermes/config.yaml:
How the CLI bridge works
The Hermes Agent CLI does not yet have a native aimlapi provider ID. Hermes Desktop bridges this by routing AI/ML API traffic through --provider custom, passing CUSTOM_BASE_URL and mapping AIMLAPI_API_KEY to OPENAI_API_KEY so the CLI can authenticate. This is transparent — from the desktop UI it behaves like any other built-in provider.
Model selection
Use exact AI/ML API model IDs. Hermes Desktop auto-discovers available models from /v1/models — no hardcoded model list.
Common examples:
openai/gpt-4oanthropic/claude-sonnet-4-5google/gemini-2.5-flash
Good starting models
openai/gpt-4ofor strong general useanthropic/claude-sonnet-4-5for coding and long-form workgoogle/gemini-2.5-flashfor lower latency and lower cost
For the full catalog, use All Model IDs.
Switch models
Open the Models screen to manage saved model configurations per provider. To change the active model mid-session, use the /model slash command in chat:
Verify the setup
After setup, open a chat and send a message. If the model responds normally, the provider is configured correctly.
You can also check the active provider and model from the chat footer or with the /status slash command.
Config checklist
Make sure these values are set:
base URL:
https://api.aimlapi.com/v1provider:
AI/ML API(shown asaimlapiin config files)model ID: exact AI/ML API model ID
API key variable:
AIMLAPI_API_KEY
Troubleshooting
AI/ML API does not appear in the provider list
The preset was added in PR #612. Update Hermes Desktop to the latest version from hermesagents.cc or the releases page.
My model is missing from the picker
The model list is pulled live from /v1/models. If a model does not appear, set it manually in ~/.hermes/config.yaml:
I get an auth error
Check that:
AIMLAPI_API_KEYis set in~/.hermes/.envthe key is valid at aimlapi.com/app/keys
no other provider is overriding the env key
Use Settings → Backup & Debug → Debug dump for a full diagnostics report.
Links
All Model IDs
Last updated
Was this helpful?