Hermes Agent - Alibaba Cloud Model Studio - Alibaba Cloud Documentation Center

Install Hermes Agent

Run the following command in your terminal. The installation script automatically installs dependencies such as Python and Git.
```
curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
```
Note
Native installation is not supported on Windows. Install WSL2 first, then run the command above in WSL2.

After installation, reload your terminal environment.

source ~/.bashrc    # If you use zsh, run: source ~/.zshrc

Verify the installation.
```
hermes --version
```
If a version number is displayed, the installation is successful.

Configure Token Plan (Team Edition)

Hermes Agent is compatible with the OpenAI protocol. Use the hermes config set command to configure the Base URL and API Key for Token Plan (Team Edition).

Obtain your Token Plan (Team Edition) API Key.

Go to the Token Plan (Team Edition) page to obtain your API Key.

Run the following commands in your terminal to configure the model provider, Base URL, API Key, and default model.

Replace YOUR_API_KEY with your actual Token Plan (Team Edition) API Key.

hermes config set model.provider custom
hermes config set model.base_url https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1
hermes config set model.api_key YOUR_API_KEY
hermes config set model.default qwen3.6-plus

The commands above write the configuration to ~/.hermes/config.yaml. You can also edit the file directly with the following content:

config.yaml example

model:
  default: qwen3.6-plus
  provider: custom
  base_url: https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1
  api_key: YOUR_API_KEY

Switch models

After configuration, use the -m flag to switch between models supported by Token Plan (Team Edition) during a conversation.

hermes chat -m qwen3.6-plus

You can also change the default model with hermes config set:

hermes config set model.default qwen3.6-plus

For a full list of supported models, see Token Plan (Team Edition) Overview.

Note

Text models (such as qwen3.6-plus and glm-5) can be used directly. Image generation models use a separate API and require integration through the extension mechanism. For more information, see Integrate multimodal generation models.

Verify the configuration

Run the following command to send a test message.

hermes chat -q "Hello"

If you receive a normal AI response, the configuration is successful.

To enter interactive conversation mode, run:

hermes

FAQ

Still connecting to OpenRouter after configuration

By default, Hermes Agent uses OpenRouter as the inference provider. To use Token Plan (Team Edition), model.provider must be set to custom. Run the following command to confirm:

hermes config set model.provider custom

For more frequently asked questions, see FAQ.