> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pre.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Messages

> POST /v1/messages in the Anthropic Messages shape, for the Anthropic SDK.

`POST https://api.pre.dev/v1/messages` accepts the Anthropic Messages request body and returns the Anthropic response. Use it with the Anthropic SDK: set the client's `baseURL` to `https://api.pre.dev` (no `/v1` — the SDK appends `/v1/messages` itself) and pass the pre.dev key as the API key. The SDK's `x-api-key` header is accepted.

Model ids are the catalog ids exactly as listed by [`GET /v1/models`](/ai-gateway/api/models), so `anthropic/claude-sonnet-5` rather than a native Anthropic model name. Any chat model in the catalog works through this shape, not only Anthropic's.

<CodeGroup>
  ```bash curl theme={null}
  curl --fail-with-body https://api.pre.dev/v1/messages \
    -H "x-api-key: $PREDEV_API_KEY" \
    -H "anthropic-version: 2023-06-01" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "anthropic/claude-sonnet-5",
      "max_tokens": 256,
      "messages": [{ "role": "user", "content": "Give me a haiku about deploys." }]
    }'
  ```

  ```typescript Node.js theme={null}
  import Anthropic from '@anthropic-ai/sdk';

  const client = new Anthropic({
    baseURL: 'https://api.pre.dev', // the Anthropic SDK appends /v1/messages itself — no /v1 here,
    apiKey: process.env.PREDEV_API_KEY!,
  });

  const message = await client.messages.create({
    model: 'anthropic/claude-sonnet-5',
    max_tokens: 256,
    messages: [{ role: 'user', content: 'Give me a haiku about deploys.' }],
  });
  console.log(message.content);
  ```

  ```python Python theme={null}
  import os
  from anthropic import Anthropic

  client = Anthropic(
      base_url="https://api.pre.dev",
      api_key=os.environ["PREDEV_API_KEY"],
  )

  message = client.messages.create(
      model="anthropic/claude-sonnet-5",
      max_tokens=256,
      messages=[{"role": "user", "content": "Give me a haiku about deploys."}],
  )
  print(message.content)
  ```
</CodeGroup>

## Streaming

Set `stream: true`, or use `client.messages.stream()` in the SDKs, for Anthropic-style server-sent events. Credits settle when the stream ends; a disconnect still charges what the model generated. See [streaming](/ai-gateway/streaming).

## Tools, system prompts, and images

Use the Anthropic `system`, `tools`, and content-block formats. The gateway forwards them to Anthropic models as-is and translates them for models from other labs.

## Headers and billing

The same `x-predev-request-id`, `x-predev-credits-charged`, and `x-predev-credits-remaining` headers are returned, and `x-predev-project-id` attributes usage. The charge is computed from `usage.cost`, the model's metered cost for the call, as on [pricing](/ai-gateway/pricing).