> ## Documentation Index
> Fetch the complete documentation index at: https://docs.venice.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Hermes Agent

> Set up Venice as the model provider in Hermes Agent for private, uncensored AI with persistent memory, custom skills, and multi-platform messaging.

[Hermes Agent](https://hermes-agent.nousresearch.com) is an open-source, self-hosted AI agent built by [Nous Research](https://nousresearch.com). It features persistent memory, autonomous skill creation, and a built-in learning loop that gets more capable the longer it runs. Point it at the Venice API and your agent gets access to 230+ models and tools across text, image, video, audio, embeddings, and more.

<Card title="Hermes Agent Docs" icon="arrow-up-right-from-square" href="https://hermes-agent.nousresearch.com/docs/">
  Full documentation, provider setup, and configuration options on the official Hermes Agent docs.
</Card>

## Why Venice + Hermes Agent?

The Venice API gives your Hermes Agent access to the full Venice platform through a single OpenAI-compatible endpoint.

| Capability        | What you get                                                     |
| ----------------- | ---------------------------------------------------------------- |
| **Text and chat** | Private and anonymized models (GLM, Qwen, Claude, GPT, and more) |
| **Image**         | Generation, editing, upscaling, and background removal           |
| **Video**         | Generation and transcription                                     |
| **Audio**         | Speech synthesis (TTS), music generation, and speech-to-text     |
| **Embeddings**    | Vector embeddings for RAG and semantic search                    |
| **Tools**         | Web scraping, web search, text parsing, and crypto RPC           |

<CardGroup cols={3}>
  <Card title="Private Inference" icon="shield-halved">
    Zero data retention. Prompts are never stored or logged
  </Card>

  <Card title="Persistent Memory" icon="brain">
    Hermes remembers context across sessions and restarts
  </Card>

  <Card title="15+ Platforms" icon="comments">
    Reach your agent on Telegram, Discord, Slack, WhatsApp, and more
  </Card>
</CardGroup>

## Setup

### 1. Install Hermes Agent

<Tabs>
  <Tab title="macOS / Linux">
    ```bash theme={"system"}
    curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
    ```
  </Tab>

  <Tab title="WSL2 (Windows)">
    Install [WSL2](https://learn.microsoft.com/en-us/windows/wsl/install) first, then run the same command:

    ```bash theme={"system"}
    curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash
    ```
  </Tab>
</Tabs>

After installation, reload your shell:

```bash theme={"system"}
source ~/.zshrc   # or: source ~/.bashrc
```

### 2. Configure Venice as your provider

Run the model setup wizard:

```bash theme={"system"}
hermes model
```

Select **Custom endpoint (self-hosted / VLLM / etc.)** and enter the following when prompted:

| Field            | Value                          |
| ---------------- | ------------------------------ |
| **API base URL** | `https://api.venice.ai/api/v1` |
| **API key**      | Your Venice API key            |
| **Model name**   | A Venice model ID (see below)  |

Get an API key from [venice.ai/settings/api](https://venice.ai/settings/api) if you don't have one yet.

<Tip>
  You can also configure Venice directly in `~/.hermes/config.yaml`:

  ```yaml theme={"system"}
  model:
    default: zai-org-glm-5
    provider: custom
    base_url: https://api.venice.ai/api/v1
    api_key: ${VENICE_API_KEY}
  ```

  And set the key in `~/.hermes/.env`:

  ```bash theme={"system"}
  VENICE_API_KEY=your-key-here
  ```
</Tip>

### 3. Pick a model

When the wizard asks for a model, choose one based on your use case:

| Use case   | Model                   | Privacy    |
| ---------- | ----------------------- | ---------- |
| General    | `zai-org-glm-5`         | Private    |
| Reasoning  | `kimi-k2-6`             | Private    |
| Coding     | `claude-opus-4-7`       | Anonymized |
| Vision     | `z-ai-glm-5v-turbo`     | Anonymized |
| Uncensored | `venice-uncensored-1-2` | Private    |

Change your model anytime with no restart needed:

```bash theme={"system"}
hermes model              # full wizard
```

Or switch mid-session:

```text theme={"system"}
/model custom:claude-opus-4-7
```

### 4. Start chatting

Open the classic CLI or the modern TUI:

```bash theme={"system"}
hermes            # classic CLI
hermes --tui      # modern TUI (recommended)
```

Try a prompt to verify everything works:

```text theme={"system"}
Summarize this repo in 5 bullets and tell me what the main entrypoint is.
```

## Connect messaging platforms

Once the CLI works, connect your messaging apps through the gateway:

```bash theme={"system"}
hermes gateway setup
```

This walks you through connecting Telegram, Discord, Slack, WhatsApp, Signal, and other platforms. Your agent becomes reachable from any connected channel, all powered by Venice.

```bash theme={"system"}
hermes gateway      # start the messaging gateway
```

## Privacy modes

Venice models in Hermes Agent follow the same [privacy tiers](/overview/privacy) as the Venice API:

* **Private** models (GLM, Qwen, DeepSeek, Llama, Venice Uncensored) run on Venice's GPU fleet. Prompts are never stored or logged.
* **Anonymized** models (Claude, GPT, Gemini, Grok) are proxied through Venice with all identifying information stripped. The third-party provider sees Venice as the customer, not you.

## Venice API skills

Hermes Agent has a built-in skills system compatible with the [Agent Skills](https://github.com/veniceai/skills) format. Venice publishes official skills that teach your agent how to use every Venice endpoint (chat, image generation, video, audio, embeddings, augment tools, and more).

Install Venice skills directly from GitHub:

```bash theme={"system"}
hermes skills install veniceai/skills
```

Or search for individual skills:

```bash theme={"system"}
hermes skills search venice
```

Hermes will discover each skill by its `SKILL.md` frontmatter and load it on demand.

## Key commands

| Command             | Description                       |
| ------------------- | --------------------------------- |
| `hermes`            | Start chatting                    |
| `hermes model`      | Change provider or model          |
| `hermes tools`      | Configure which tools are enabled |
| `hermes gateway`    | Start the messaging gateway       |
| `hermes --continue` | Resume your last session          |
| `hermes doctor`     | Diagnose issues                   |
| `hermes update`     | Update to the latest version      |

## Resources

<CardGroup cols={2}>
  <Card title="Hermes Agent Docs" icon="book" href="https://hermes-agent.nousresearch.com/docs/">
    Official documentation
  </Card>

  <Card title="GitHub" icon="github" href="https://github.com/NousResearch/hermes-agent">
    Source code and releases
  </Card>

  <Card title="Venice Model Catalog" icon="list" href="/models/text">
    Browse available models
  </Card>

  <Card title="Venice Privacy" icon="shield-halved" href="/overview/privacy">
    How Venice protects your data
  </Card>
</CardGroup>