Aetherfy Documentation
Aetherfy is two products behind one control plane, one account, and one API key:
- Vector database — a Qdrant-compatible vector API, reached through the Python SDK, the JS/TS SDK, or plain HTTP.
- Agent compute — push-to-deploy agents that run as long-lived services, or as tasks that run once and exit, optionally on a recurring schedule.
You can use either product on its own. An agent running on Aetherfy can read and write an Aetherfy collection, but nothing requires it to.
Start here
Pick the Aetherfy product you want to use first — each path ends with something working, not with more reading.
- Vector database quickstart — get an API key, install an SDK, create a collection, upsert points, and get search results back. Covers both Python and JS/TS.
- Deploy an agent — push a repository to Aetherfy and have it run as a service or as a scheduled task.
Both quickstarts assume nothing beyond an Aetherfy account, which you can create at app.aetherfy.com/login with email or GitHub.
Where Aetherfy runs
Aetherfy runs in three regions. These AWS-style codes are exactly what the
Aetherfy API, both SDKs, and aetherfy.yaml accept:
| Code | Location |
|---|---|
us-east-1 | US East (Virginia) |
eu-central-1 | EU Central (Frankfurt) |
ap-southeast-1 | Asia Pacific (Singapore) |
How many of those you get depends on your plan, and this is the single most common misreading of Aetherfy:
- Free and Starter are single-region. The region is chosen by your first resource, and everything you create afterwards lives there. Data is not replicated to the other two regions on these plans.
- Performance and above are multi-region. A resource can be placed in up to 3 regions, with its data replicated between them.
So a request does not automatically reach your data from anywhere — on Free and Starter it reaches the one region your resources actually live in. Plan details are at aetherfy.com/pricing , and the mechanics of placement and replication are in Regions & replication.
Map of the Aetherfy documentation
Everything Aetherfy documents lives under one of five sections.
Vector database — the Qdrant-compatible side of Aetherfy. SDK reference for the Python and JS/TS clients, REST API reference for plain HTTP, Filtering for narrowing a search by payload, Search tuning, Limits, and Errors.
Agent compute — deploying code to Aetherfy. Deploy an agent, aetherfy.yaml reference, Scheduled tasks for work that runs on a recurring schedule, Runs & logs, Secrets, and GitHub integration.
CLI — afy, the Aetherfy command-line tool: installation and
authentication, then per-command pages for deploying, tailing logs, and managing
agents, secrets, and workspaces.
Platform — the parts of Aetherfy shared by both products: Regions & replication, API keys, Workspaces, Billing & spend caps, Limits, and Using AI with Aetherfy.
Examples — a cookbook of complete, runnable tasks, each written in both Python and JS/TS.
Using these Aetherfy docs with an LLM
These pages are written to survive being retrieved one chunk at a time: each page states its own context, each code sample is complete and runnable, and no section depends on the one before it. That makes them usable as source material for a coding assistant working against Aetherfy.
If you want to point an LLM, an editor, or an MCP client at this documentation, Using AI with Aetherfy explains how.
Getting to your Aetherfy account
The Aetherfy dashboard is where you manage keys, usage, and deployed agents.
- Sign in: app.aetherfy.com/login
- Dashboard home: app.aetherfy.com/dashboard
- API keys: app.aetherfy.com/dashboard/settings/api-keys