Providers & integrations
Hermes One sits in front of the Hermes Agent, so it supports the providers and integrations the agent does, with GUI setup for each.
LLM providers
The first-run picker mirrors the agent's native canonical providers; any OpenAI-compatible endpoint routes through the Local presets.
| Provider | Notes |
|---|---|
| OpenRouter | 200+ models via a single API (recommended) |
| Anthropic | Direct Claude access |
| OpenAI | Direct GPT access |
| Google (Gemini) | Google AI Studio |
| xAI (Grok) | Grok models |
| Nous Portal | Free tier available |
| Qwen | QwenAI models |
| MiniMax | Global and China endpoints |
| Hugging Face | 20+ open models via HF Inference |
| Groq | Fast inference (also used for voice/STT) |
| Atlas Cloud | OpenAI-compatible gateway (sponsor) — DeepSeek, Qwen, GLM, Kimi, MiniMax… base URL pre-configured |
| Local / Custom | Any OpenAI-compatible endpoint |
Local model presets
Bundled presets that need no API key — but the server must already be running:
- LM Studio
- Atomic Chat
- Ollama
- vLLM
- llama.cpp
Select Local LLM and either pick a preset or supply a custom OpenAI-compatible base URL.
Messaging platforms (gateways)
Connect from the Gateway screen so you can talk to your agent from your existing chat apps. 16 supported:
Telegram · Discord · Slack · WhatsApp · Signal · Matrix/Element · Mattermost · Email (IMAP/SMTP) · SMS (Twilio & Vonage) · iMessage (BlueBubbles) · DingTalk · Feishu/Lark · WeCom · WeChat (iLink Bot) · Webhooks · Home Assistant.
See Features → Gateway.
Tool integrations
External services the agent's toolsets can use:
Exa Search · Parallel API · Tavily · Firecrawl · FAL.ai (image generation) · Honcho · Browserbase · Weights & Biases · Tinker.
Memory providers
Discoverable memory backends configurable from the Memory screen:
Honcho · Hindsight · Mem0 · RetainDB · Supermemory · ByteRover.
Speech-to-text
Voice transcription is routed through the Hermes API server
(/api/audio/transcribe), independent of the selected chat model — so local
Whisper, Groq, OpenAI, ElevenLabs, and command/plugin STT providers all work
regardless of which model you're chatting with.
Where credentials go
Provider API keys are stored via the Hermes secrets provider — by default
~/.hermes/.env, or an opt-in command/vault helper. Local providers need no key.
See Configuration → Secrets.