Early access, by application.
Vesta is in early access. We're working with a small number of design partners before any self-serve plan opens.
Early access
For teams running an MCP server or agent-facing surface in production who want to know how agents use it. You get the analysis directly, and a strong say in what Vesta becomes.
The trade is access to your anonymised trace patterns, agreed in writing. No raw data leaves your control.
- ✔ Vesta on your real production traffic
- ✔ Recommendations from a data scientist, not a dashboard
- ✔ Deploy in your own cloud (Azure / GCP / VPC) if you need to
- ✔ Shared Slack or Teams channel
- ✔ A direct line to the person building it
Vesta Cloud
A self-serve version for smaller teams. Install the SDK, see your surface, get recommendations. Not open yet.
Pricing lands once early access tells us what the product should be.
- • Install the SDK and add one call
- • Usage, sequences and failure patterns
- • Automated recommendations on your surface
- • End-customer analytics once you supply session context
Questions worth asking
Will the SDK slow my server down?
No. It wraps your tools/call handler and exports asynchronously over OpenTelemetry. It's fail-open: if Vesta breaks, your server keeps serving.
Can Vesta run inside our own cloud?
Yes. For design partners we deploy the whole pipeline inside your tenant, Azure or GCP. Your traces never leave your environment. The only thing that crosses between us is a container image you pull.
Which MCP servers do you support?
Python — both the official mcp SDK and FastMCP — and TypeScript on the official MCP SDK. OpenTelemetry underneath, over stdio, streamable HTTP or SSE.
Is this observability?
No. Vesta is analytical, not operational. It runs on accumulated traffic and produces periodic recommendations about your surface, not live alerts. It sits on top of your logs and traces, it doesn't replace them.