Deploy Managed Agents. One click. Done.
Hive is SAQI/AI's Managed Agents platform. Point it at Hetzner, Vultr or our Proxmox Servers and it deploys a full agent runtime — multi-agent management, skills marketplace, channel integrations, and a browser-based playground — in under a minute. Your VPS, your data, your root.
Agent infrastructure, preinstalled.
Every Hive deployment ships with the full agent stack. No Dockerfiles. No reverse-proxy config. No TLS scripts.
Multi-agent runtime
Create, configure and run multiple agents per deployment. Each agent has its own skills, env vars and memory.
Agent playground
Chat with any deployed agent from the dashboard. Inspect tool calls. Iterate without redeploying.
Skills marketplace
Browse, install, update and remove skills from the marketplace — package-manager ergonomics for agent capabilities.
Channel integrations
Wire agents to Slack, Discord, Telegram and WhatsApp with app tokens. No glue code.
SSH key management
Add ed25519 or RSA keys once. Deploy to any instance passwordlessly. Rotate and revoke from the console.
Export as image
Snapshot a configured deployment and redeploy it elsewhere. Portability is first-class, not an afterthought.
Environment isolation
Per-agent environment variables and secrets, scoped and encrypted. Audit what was set and when.
i18n out of the box
Console in English, French, Spanish and German. Agents speak whatever you configure.
Open source core
The underlying runtime is open source. No vendor lock-in: export your deployment and run it anywhere.
Flat monthly. No surprises.
You pay for the Hive deployment layer. Your VPS bill is separate and goes to your VPS provider. Cancel any month.
Starter
- ▸1 deployment
- ▸Up to 3 agents
- ▸Community skills
- ▸1 region
- ▸Community support
Pro
- ▸5 deployments
- ▸Unlimited agents
- ▸All channel integrations
- ▸Global regions
- ▸Priority email support
- ▸$60 in AI model credits
Scale
- ▸50 deployments (hard limit)
- ▸SSO + audit logs
- ▸All channels + webhooks
- ▸$180 in AI model credits
- ▸SLA-backed support
Enterprise
- ▸Dedicated deployments
- ▸Custom SLA
- ▸DPA + procurement
- ▸Named support engineer
- ▸Volume AI credits
Prices are in USD. AI model credits apply to Hive's built-in model routing when enabled; bring-your-own-key is free and unlimited.
Answers.
- Hive is SAQI/AI's Managed Agents platform — a deployment and management layer that installs a full agent runtime on Hetzner, Vultr or our Proxmox Servers. One click in the console and your VPS is ready to run AI agents.
- A deployment is a single agent runtime instance running on one of your VPSes. Each deployment can host multiple agents, each with its own skills, memory and env vars. Plans are priced per deployment, not per agent.
- Yes. Hive deploys to Hetzner, Vultr or our Proxmox Servers. You keep root and your infrastructure bill stays with your provider.
- No. The underlying runtime is open source. Every deployment can be exported as an image and redeployed outside Hive at any time. You own the runtime.
- Pro and Scale include $60 and $180 in AI model credits respectively, usable against Hive's routed models. You can also bring your own API keys (OpenAI, Anthropic, local models, etc.) and pay those providers directly.
- Yes — Enterprise plans remove the cap. Reach out via the contact form and we'll scope it.
Ship an agent on your own infrastructure.
Hive is live today. Agents, Inference, and GPU are in private beta — request access and we'll reach out.