# infrai

> infrai gives your app 14 production backend modules — AI, video,
> email, SMS, scheduling, storage, payments and more — behind one SDK, one API key,
> one wallet and one bill. AI inference is billed at China 0% /
> Western 5% markup. Every response carries cost/vendor/latency
> metadata; every error carries a machine-readable hint and a retryable flag.

The first line of any infrai program:

```python
from infrai import ai, infra, account
```

## 14 GA modules (93 API routes)

- `ai-runtime` (/v1/ai) — ai.chat, ai.embed, ai.image
- `ai-video` (/v1/video) — video.generate, video.status, video.cancel
- `comm-email` (/v1/email) — email.send, email.get, email.list
- `sms` (/v1/sms) — sms.send, sms.otp, sms.verify
- `scheduling` (/v1/scheduling) — scheduling.cron.create, scheduling.cron.list, scheduling.queue.publish
- `observability` (/v1/observability) — observability.error.capture, observability.event.track, observability.span.report
- `public-url` (/v1/public-url) — public_url.create, public_url.claim, public_url.domain.create
- `captcha` (/v1/captcha) — captcha.verify, captcha.widget.create
- `pdf` (/v1/pdf) — pdf.generate, pdf.merge, pdf.split
- `image-process` (/v1/image) — image.process, image.metadata
- `realtime` (/v1/realtime) — realtime.token.issue, realtime.channel.create, realtime.publish
- `storage` (/v1/storage) — storage.bucket.create, storage.bucket.list, storage.object.presign
- `analytics` (/v1/analytics) — analytics.track, analytics.identify, analytics.funnel
- `billing` (/v1/billing) — billing.charge.create, billing.charge.get, billing.refund.create

## SDK install

- Python: `pip install infrai`
- TypeScript: `npm install @infrai/sdk`
- Go: `go get github.com/infrailab/infrai-go`
- Rust: `cargo add infrai`
- Java: `io.infrailab:infrai-sdk:0.2.0`
- C#: `dotnet add package Infrai`
- Ruby: `gem install infrai`
- PHP: `composer require infrailab/infrai`

## API hosts

- Western: https://api.infrailab.com
- China: https://api.cn.infrailab.com

## For AI agents & tools (machine discovery)

If you are an AI coding assistant or agent, integrate infrai directly:

- MCP server (one command, all 14 modules as tools): `npx @infrai/mcp-server`
  Works with Claude Desktop, Claude Code, Cursor, Cline, Continue, Windsurf, Aider.
- OpenAPI 3.1 spec (generated from SSOT, with request schemas + examples):
  https://infrailab.com/.well-known/openapi.json
- Discovery manifest (full capability table: method/path/params/examples/errors):
  https://infrailab.com/.well-known/infrai-discovery.json
- MCP registry descriptor: https://infrailab.com/.well-known/mcp.json

CLI self-discovery (the `infrai` command ships with the Python SDK):

```bash
infrai modules list                 # capabilities available in your region
infrai params <capability>           # JSON Schema for a capability (e.g. ai.chat)
infrai examples <capability>         # runnable request examples
infrai help <capability>             # signature + params + examples
```

## Links

- Site: https://infrailab.com
- Docs: https://infrailab.com/docs