# Astrolabe Cloud

> Hosted AI routing, billing, observability, and stack control for production AI APIs.

## Docs

- [Chat Completions](https://docs.astrolabe.run/api-reference/endpoint/chat-completions.md): Compatibility endpoint for OpenAI Chat Completions clients. Uses the same Cloud stack, billing, and trace runtime as /v1/responses.
- [Responses](https://docs.astrolabe.run/api-reference/endpoint/responses.md): Primary Astrolabe Cloud inference endpoint. Supports virtual models, concrete model IDs, routing stack selection, streaming, billing, and route traces.
- [API Introduction](https://docs.astrolabe.run/api-reference/introduction.md): Astrolabe Cloud gateway endpoints, authentication, headers, and error model.
- [Configuration](https://docs.astrolabe.run/configuration.md): Configure Astrolabe Cloud through workspace keys, billing, stack assignment, and custom stack policy.
- [Hosted Gateway](https://docs.astrolabe.run/hosted-gateway.md): Use Astrolabe Cloud's hosted API gateway for production AI traffic.
- [How Astrolabe Cloud Works](https://docs.astrolabe.run/how-it-works.md): Hosted request lifecycle, stack resolution, billing, fallback, and traceability.
- [Astrolabe Cloud](https://docs.astrolabe.run/index.md): Hosted AI routing with workspace keys, prepaid billing, routing stacks, and request-level traceability.
- [Model Roster](https://docs.astrolabe.run/model-roster.md): Public virtual models and hosted model roles available through Astrolabe Cloud.
- [OpenClaw Integration](https://docs.astrolabe.run/openclaw.md): Use Astrolabe Cloud as a hosted provider for OpenClaw and other OpenAI-compatible clients.
- [Quickstart](https://docs.astrolabe.run/quickstart.md): Create a Cloud key, fund your workspace, and send your first request through Astrolabe Cloud.
- [Routing Reference](https://docs.astrolabe.run/routing-reference.md): Stack resolution, model selection, fallback, verification, and trace metadata.
- [Routing Stack Concepts](https://docs.astrolabe.run/routing-stacks/concepts.md): How Astrolabe Cloud uses managed and custom stacks to choose models.
- [Customize Stacks](https://docs.astrolabe.run/routing-stacks/customization.md): Configure model boundaries, objectives, fallback, verification, budget, memory, and observability.
- [Request Stack Selection](https://docs.astrolabe.run/routing-stacks/request-selection.md): Select a specific routing stack and set request-level routing preferences.
- [Troubleshooting](https://docs.astrolabe.run/troubleshooting.md): Diagnose Cloud auth, billing, stack resolution, policy blocks, rate limits, and unbillable requests.

## OpenAPI Specs

- [openapi](https://docs.astrolabe.run/api-reference/openapi.json)

## Optional

- [Astrolabe](https://astrolabe.run)
- [Cloud app](https://app.astrolabe.run)