Beta program open
Meet Hermes — your personal AI agent, always on.
Deploy a managed Hermes agent in one click. Connect it to Telegram, WhatsApp, and Discord, and let it work 24/7 on 10M+ free tokens. No coding required.
Dockerized 1-click setup.
Powered by SNBD API.
Swap the base URL, done.
Full data sovereignty.
Exclusive offer
Powered by the Tonic API system.
Every Hermes plan ships with a generous free token bundle across our best models — tuned for reasoning, coding, and everyday tasks. Enough to run heavy, agentic workloads across multiple channels without surprise API bills.
Real-world use cases
What can you do with Hermes?
Round-the-clock community & customer support.
Wire Hermes to Discord or Slack to answer FAQs, resolve tickets, and route escalations — instantly, in 50+ languages.
- Auto-answer common questions
- Ticket triage & routing
- Multilingual replies
Simple as 1-2-3
Three steps to your own AI agent.
No command line, no config files, no headaches.
Choose from Gemma, Llama 3, Mistral, or the included SNBD API. Swap anytime.
Pick Spark, Messenger, Oracle, or Olympus — scale up whenever you outgrow it.
Wire Hermes to Telegram, Discord, or Slack — or use the OpenAI-compatible API.
Everything you need
Key features, built for real work.
Dockerized deploys, zero server tinkering.
Slack, Discord, Telegram — and WhatsApp on higher tiers.
Swap OpenAI keys with a Hermes key and go.
Monitor usage, tweak settings, rotate keys from anywhere.
Gemma, Llama, Mistral, DeepSeek — one click each.
Your prompts and memory stay on your VPS.
For developers
A drop-in for the OpenAI SDK.
Same API surface, same SDKs. Point base_url at your Hermes endpoint and your existing apps just work.
from openai import OpenAI
client = OpenAI(
base_url="https://api.hermes.hosttonic.app/v1",
api_key="ht_sk_•••"
)
reply = client.chat.completions.create(
model="gemma-3-reasoning",
messages=[{"role": "user", "content": "Plan my week."}]
)
print(reply.choices[0].message.content)Pricing
Pick your Hermes plan.
Every plan bundles free SNBD tokens, 1-click deploy, and a 30-day money-back guarantee.
Currency
Choose billing period
Spark
Kick off your personal AI agent — perfect for solo makers.
৳799/mo · save 20%
৳15,336 billed every 24 months
Choose plan- 1 vCPU · 2 GB RAM · 40 GB NVMe
- 3 TB bandwidth
- 1-click Hermes agent deploy
- 10M free SNBD tokens / month
- Telegram · Discord · Slack connectors
- OpenAI-compatible API endpoint
- Free SSL · Weekly backups
Messenger
The sweet spot for always-on assistants and small teams.
৳1,399/mo · save 20%
৳26,856 billed every 24 months
Choose plan- 2 vCPU · 4 GB RAM · 80 GB NVMe
- 6 TB bandwidth
- 25M SNBD tokens / month
- Gemma · Llama 3 · Mistral model swap
- Persistent memory & scheduled runs
- Multi-channel bots (WhatsApp add-on)
- Daily backups · Snapshots
Oracle
For agentic workflows, coding copilots, and research bots.
৳2,599/mo · save 20%
৳49,896 billed every 24 months
Choose plan- 4 vCPU · 8 GB RAM · 160 GB NVMe
- 12 TB bandwidth
- 60M SNBD tokens / month
- Sub-agent orchestration
- Vector DB (pgvector) preinstalled
- Priority model routing
- 24/7 engineer support
Olympus
Enterprise-grade Hermes fleet for production automations.
৳4,999/mo · save 20%
৳95,976 billed every 24 months
Choose plan- 8 vCPU · 16 GB RAM · 320 GB NVMe
- 24 TB bandwidth
- Unlimited SNBD tokens (fair use)
- Local Ollama & fine-tuned models
- Dedicated IP · VPC isolation
- 99.9% uptime SLA
- Hands-on onboarding
All plans include a 30-day money-back guarantee. 1 USD = 125 BDT · 1 EUR = 155 BDT.
FAQ
Questions about Hermes.
What is Hermes?+
Hermes is Host Tonic's managed AI-agent runtime. You deploy an always-on assistant in one click, connect it to messaging apps, and drive it via chat or an OpenAI-compatible API.
Do I need to bring my own API keys?+
No. Every plan includes free SNBD tokens across Gemma, Llama, Mistral and more. You can also plug in your own OpenAI, Anthropic, or Gemini keys anytime.
What models are available?+
Gemma by Google DeepMind, Llama 3, Mistral, DeepSeek, Qwen, plus the SNBD API. Model swap is a single click in your dashboard.
Which messaging apps are supported?+
Telegram, Discord, and Slack on every plan. WhatsApp is available from the Messenger tier upward.
Is Hermes really OpenAI-compatible?+
Yes — same endpoints and SDKs. Point base_url to your Hermes URL, drop in your Hermes API key, and existing apps like Cursor, LibreChat, or your custom code just work.
Where is my data stored?+
On your Host Tonic VPS. Prompts, memory, and agent files never leave your instance. Backups and snapshots stay on the same infrastructure.
Ready to start?
Supercharge your workflow with 10M free tokens.
Join the Hermes beta and deploy your personal AI agent in 60 seconds — fast, open, and built for real work.