Deploy AI agents.Not infrastructure.
Mantlecore is the cloud built for open-source AI agents. One click boots your agent's machine — VM-isolated, persistent storage, its own subdomain. No servers. No Docker. No setup.
Start Deploying7-day free trial · No charge today · Cancel anytime
Pick Your AI Agent
Agent Zero
Hermes AgentChoose Your AI Model
Plan
2 vCPU / 4 GB
AI Model
GPT ModelsStart Working
Agent1
Works with leading AI agents
The Problem
AI frameworks are powerful.
Getting them running shouldn't take a weekend.
What you get with chat AI
You ask for outcomes. You get suggestions.
What it takes to self-host an AI framework
- 1.Provision a VPS$20–200/mo
- 2.Install Docker, runtime, GPU drivers~2 hrs
- 3.Wire up reverse proxy + TLS~1 hr
- 4.Lock down firewall + SSH~1 hr
- 5.Set up backups + monitoring~2 hrs
- 6.Patch CVEs, rotate keys, on-callforever
The most powerful AI frameworks are locked behind a weekend of plumbing.
Powerful AI frameworks exist. The gap is the weekend you’d spend on Docker, TLS, and CVE patching before you write a single agent prompt.
Understand the Difference
There's a difference between AI that talks
and AI that works.
ChatGPT answers questions. An AI agent does your work.
| Capability | ChatGPT | AI Agent on MantlecoreBetter |
|---|---|---|
| What it does | Answers your questions | Does your work |
| Memory | Forgets between chats | Remembers everything |
| Browse the web | Sometimes, unreliably | Yes, with full access |
| Run code | In a sandbox, then forgets | Persistent environment |
| Use tools | Limited | Browser, files, APIs — all connected |
| Multi-step tasks | You guide every step | Works autonomously, reports back |
| Your data | Sent to OpenAI servers | Stays in your private workspace |
| AI model choice | Locked to OpenAI | Any model: OpenAI, Google, Anthropic, more |
How It Works
3 steps. Zero infra.
Live in seconds.
No Docker. No YAML. No reverse proxy. API keys optional.
Pick your framework
Choose your framework and its model from a dropdown — any provider, or bring your own key.
Framework
Agent Zero
HermesDefault Model
Click deploy
One click. Your framework boots in an isolated machine with a real filesystem, persistent storage, and its own subdomain. Live in seconds.
My Agent
Start working
Give it a task in plain English. Your agent researches, codes, browses, builds tools, and delivers results. Not suggestions — results.
Use Cases
What will you delegate first?
Real tasks that real people delegate to their AI agent every day.
Deep Research
Your agent browses the web, reads papers, cross-references sources, and compiles structured reports — turning hours of research into minutes.
>“Research the top 10 competitors in the AI agent space. Compare pricing, features, funding, and market positioning. Create a detailed report.”
Data Analysis
Upload a dataset. Your agent cleans the data, runs statistical analysis, builds visualizations, and writes an executive summary with actionable insights.
>“Here is 12 months of sales data. Find trends, seasonal patterns, and top products. Create charts and a summary report with recommendations.”
Writing & Content Creation
Your agent researches topics, analyzes top-ranking content, and drafts blog posts, newsletters, or social media copy — optimized for your audience and voice.
>“Research trending topics in AI automation. Draft a 2,000-word blog post with SEO keywords, internal links, and a compelling hook.”
Software Development
Your agent writes code, sets up projects, installs dependencies, runs tests, and debugs issues — a full-stack developer that never sleeps.
>“Scaffold a FastAPI service with JWT auth, SQLAlchemy models, and CRUD endpoints. Include a Dockerfile and pytest suite.”
Stock & Market Analysis
Your agent pulls market data, calculates technical indicators, analyzes earnings reports, and builds watchlists with buy/sell signals and risk assessments.
>“Pull NVDA, AAPL, MSFT financials. Compute RSI + MACD, compare P/E ratios, and output a buy/hold/sell summary as JSON.”
Workflow Automation
Your agent connects services, moves data between APIs, sends notifications, and handles repetitive tasks — wired up in plain English instead of YAML.
>“Every morning, pull new GitHub issues labeled 'bug', enrich them with linked PRs, and post a digest to #engineering on Slack.”
Connectors
Give your agent your tools.
Never your passwords.
One OAuth flow per account — tools appear in every agent you allow, natively over MCP. No API keys in env vars, no token rotation, no glue code. Access tokens live server-side; the workspace never sees them.
>create_issueissue #142 opened
Research Agent
Agent Zero
Support Agent
Hermes
one connector · apply selectively per agent
Security & Trust
Your own private AI workspace.
Completely isolated. Completely yours.
VM-isolated by default. One click to deploy.
A sealed perimeter only your traffic crosses
Built to be unbreakable
Your agent can never be broken.
Instant spawn, full-state snapshots, auto-snapshot on update, and daily off-site backups. Every operation is reversible — your agents are unbreakable by construction.
No Lock-In
Works with every AI model.
Locked to none.
Use your own API key or let us handle it. Switch AI models anytime — no lock-in, ever. New accounts start with $2 of free AI credit to try it out.
How We Compare
Not all AI is created equal.
| Capability | Chat AI | Manus | MantlecoreOpen |
|---|---|---|---|
| Works independently | Every step needs you | Yes, but locked to their system | Yes — pick any AI agent |
| Choice of AI agent | One size fits all | Their agent only |  Agent Zero |
| Your data stays private | Sent to provider | Sent to Manus servers | Stays in your private workspace |
| Choose your AI model | Locked to one provider | Platform picks for you | Any model — or we pick for you |
| Remembers your work | Forgets between sessions | Per-task only | Remembers everything |
| Runs real software | Limited sandbox | Cloud sandbox | Full workspace with real tools |
| Pricing | $20/mo subscription | Opaque per-task credits | Predictable monthly plans |
| Security | Shared infrastructure | Shared infrastructure | Your own private workspace |
The gap between wanting AI and using AI — that's where your time goes.
Get StartedPricing
Your AI agent. Your budget.
Every plan includes all AI agents, all features, and your own private workspace. The only difference is how much power you get. New accounts start on Starter with a 7-day free trial — no charge today.
Starter
7-day free trialEverything you need to get started.
Perfect for personal projects, learning, and automating everyday tasks.
2
vCPU
4
GB RAM
20
GB SSD
- All AI agents included
- 20 GB SSD storage included
- Web-based interface
- Expandable storage add-ons
- Daily off-site backups (7-day history)
- 1 manual snapshot
- Private workspace security
- $2 free AI credit to start
Builder
More power for serious work.
For professionals running research, analysis, and complex automations.
4
vCPU
8
GB RAM
30
GB SSD
- Everything in Starter, plus:
- 2x CPU and memory
- 30 GB SSD storage included
- Expandable storage add-ons
- 30-day backup history
- 3 manual snapshots
- Priority email support
- $2 free AI credit to start
Pro
Maximum power. Zero compromise.
For professionals with heavy workloads that need sustained performance.
8
vCPU
16
GB RAM
50
GB SSD
- Everything in Builder, plus:
- 4x CPU and memory
- 50 GB SSD storage included
- Expandable storage add-ons
- 30-day backup history
- 5 manual snapshots
- Dedicated support
- $2 free AI credit to start
Every plan includes: All AI agents · Managed AI models or use your own key · Private workspace security · Daily backups · One-click restore · Web-based interface · Persistent storage
Frequently Asked Questions
Your AI framework is one click away.
Stop configuring servers. Deploy Agent Zero or Hermes — VM-isolated, zero config, running in seconds.
7-day free trial · No charge today · Cancel anytime
Stay in the loop
Get notified about product updates, new AI agents, and automation tips.
Unsubscribe anytime · No spam, ever