Skip to content
Documentation menu
API, SDKs & AI assistants
Back to section

Choose how to manage your services

Connect an AI assistant, use the command line, or build an integration with an SDK.

Use the portal, an AI assistant, or API tools for scripts and applications.

Current API, CLI and MCP release

SDK/CLI 0.1.17 and contract 2026-10-07.1 document 182 customer API operations. Download the immutable Python, Node.js, PHP or Go archives and verify SHA256SUMS. The Python archive includes both bf and bf-api, plus the local bf-api-mcp server.

Start with bf version, bf capabilities and bf regions. Production uses https://api.blazingfast.io and download links in the installation guide. Configure the origin without /v1.

CLI lifecycle commands generally make write keys optional; DNSSEC enable/disable require an explicit --request-key. Where optional: the CLI saves and prints a generated key before sending when omitted. Preserve that key for explicit retries; inspect an unknown outcome before retrying. Read requests do not create write receipts.

Hosted MCP uses https://mcp.blazingfast.io/mcp and OAuth with your selected scopes and connection policy. Local MCP uses your API key and exposes read-only tools by default. Existing OAuth connections need fresh consent for additional scopes.

Read the full endpoint map and use GET /v1/version plus GET /v1/capabilities to check deployed availability. Check GET /v1/version and GET /v1/capabilities before using optional workflow features. Strict response validation is enforced. Use capability discovery and your account permissions to check which workflows are available.

Current purchase and operation workflows

SDK/CLI 0.1.17 supports contract 2026-10-07.1. Read GET /v1/version and GET /v1/capabilities, then discover product purchase policies and requirements before configuring a purchase. A quote is optional: discovery can lead directly to an authorized deployment. Quotes do not reserve stock or lock prices; send an authorized max_total ceiling for the full final order amount.

Draft order creation does not charge the account. Atomic deployment is the documented exception: it combines order creation, payment of the full final order amount from account balance and queued provisioning. Paid is not ready; accepted is not completed. Service lifecycle status is separate from runtime power state.

Supported tracked deployment, VPS and dedicated writes can return a canonical operation receipt and be recovered using their original idempotency key. This does not extend recovery to every write or historical request. See HTTP, SDK, CLI and MCP examples. DNSSEC uses its separate endpoint and state contract.

Choose a starting point

What you want to do Start here What you need
Ask ChatGPT about your services Connect an AI assistant Your BlazingFast account
Run commands from a terminal Install an SDK or the CLI Python 3.10+ and an API key
Call the API from your application Install an SDK Python, Node.js, PHP or Go and an API key
Build a custom HTTP client API reference API authentication and HTTP experience

MCP is the connection protocol that lets an AI assistant use BlazingFast tools. The hosted connection uses your BlazingFast sign-in. The local MCP server is installed on your computer and uses an API key.

CLI names

Use bf for new integrations. bf-api is an actively supported compatibility alias of the same CLI, not a separate lower-level interface. Both are included in the Python SDK package and use the same commands, authentication and version.

Make your first read

For an AI connection, ask: “Show my services, their status, IP addresses and paid-through dates.”

For the CLI, run bf service list after following the installation guide. This lists services without changing them. Select the returned service ID when a command needs one; do not guess an ID.

What you can manage

Available actions depend on the services you own and the permissions you grant. They include billing and invoices, VPS and dedicated servers, DNS and Website Protection, domains, TCP Proxy, public SSH keys, firewall policies, and storage volumes.

Start with viewing permissions. Add the specific change permissions your workflow needs in API keys or AI connections.

Continue with a task