Pro-16: 16 GB high-memory VPS
16 GB NVMe VPS for a single AI agent or a small crew that runs around the clock.
Who it's for
The entry point into the Pro range. 16 GB is comfortable for one always-on agent, a staging environment, or a small CrewAI/LangGraph crew that calls an LLM API rather than running the model locally. If your workload is mostly outbound API calls plus some local state, this is the sweet spot.
Typical uses
- ▸A single autonomous agent running 24/7 (trading bot, monitor, scraper)
- ▸Staging/dev box for testing agent pipelines
- ▸Small multi-agent crew (2–4 agents) hitting OpenAI/Anthropic APIs
- ▸Discord/Telegram bot with a local database
FAQ
Is 16 GB enough for a CrewAI crew?
Yes for a small crew (a handful of agents) — the heavy lifting is the LLM API, not the box. If you keep long conversation histories in memory or add a vector store, look at Pro-32 or Pro-48.
Do I need a GPU?
No. Pro plans run Python and make API calls to your LLM provider; the model runs on their GPUs, not yours. CPU + RAM is what matters here.
Can I pay in crypto without KYC?
Yes — USDC or USDT on Base, Ethereum or Polygon, email signup only, no identity check.