> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.milliseconds.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.milliseconds.ai/_mcp/server.

# Overview

milliseconds.ai is an API for small models that do one job each, fast and at a low price. One API key calls every model. The same credits pay for every model.

## The models

| Model                                     | Job                                                                                              | Endpoint                                   | Price                                               |
| ----------------------------------------- | ------------------------------------------------------------------------------------------------ | ------------------------------------------ | --------------------------------------------------- |
| [decision-machine-1](/decision-machine-1) | Typed decisions over text or an image: yes/no, classify, rate, answer, extract, entities, verify | `POST /v1/decision-machine-1/{capability}` | \$0.04 per million input tokens                     |
| [rakeaudio](/rakeaudio/overview)          | Audio transcription with word timestamps and speaker labels                                      | `POST /v1/rakeaudio/transcribe`            | \$0.12 per audio hour with speakers, \$0.10 without |

Both models answer in one synchronous call. There is no job to poll and no webhook to receive.

## One base URL, one key

Every call goes to one base URL.

```
https://api.milliseconds.ai
```

Every request needs an API key from the [console](https://console.milliseconds.ai/signup). Send it as `Authorization: Bearer test_sk-...`. See [Authentication](/authentication).

Test keys include 125 million free input tokens each month. Production keys use purchased credits. Every model bills in input tokens at \$0.04 per million, so one balance covers both. See [Pricing](/reference/pricing).

## One call to each model

**`decision-machine-1`**

```bash title="decision-machine-1"
curl -X POST https://api.milliseconds.ai/v1/decision-machine-1/classify \
  -H "Content-Type: application/json" \
  -H "authorization: Bearer test_sk-..." \
  -d '{
    "text": "I was charged twice for my subscription this month and support has not replied.",
    "labels": {
      "billing": "payments, invoices, charges, refunds",
      "shipping": "delivery, tracking, returns",
      "account": "login, password, profile settings"
    }
  }'
```

**`rakeaudio`**

```bash title="rakeaudio"
curl -X POST "https://api.milliseconds.ai/v1/rakeaudio/transcribe?diarize=true" \
  -H "authorization: Bearer test_sk-..." \
  -H "Content-Type: audio/wav" \
  --data-binary @call.wav
```

The [Quickstart](/quickstart) shows both responses.

#### [decision-machine-1](/decision-machine-1)

Eight typed decisions, an OpenAI-compatible surface and a TypeSafe-compatible surface.

#### [rakeaudio](/rakeaudio/overview)

Transcripts with word timestamps and up to four speakers, in 25 European languages.

## Next

* [Quickstart](/quickstart) — your first call to each model.
* [Authentication](/authentication) — test keys and production keys.
* [Limits and rate limits](/reference/limits) — the limits every model shares.