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

# Claude Haiku 5.5

> Use Claude Haiku 5.5 through AnyFast for fast classification, extraction, routing, and agent tasks.

Claude Haiku 5.5 is Anthropic's fast model for high-volume, latency-sensitive work. Call it through AnyFast's Anthropic Messages API at `POST /v1/messages` using the model ID `claude-haiku-5-5`.

## Key capabilities

* **1M context window** — Supports up to 1M context tokens and 128K output tokens according to Anthropic's model specification
* **Adaptive thinking** — On by default; use `output_config.effort` to balance speed and depth
* **Fast routine work** — Suited to classification, extraction, routing, and high-volume assistants
* **Tool use and vision** — Accepts text and image inputs and can return text or tool calls
* **Model ID** — Use `claude-haiku-5-5` in requests

## Quick example

<CodeGroup>
  ```bash cURL theme={null}
  curl https://www.anyfast.ai/v1/messages \
    -H "Authorization: Bearer YOUR_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "claude-haiku-5-5",
      "max_tokens": 1024,
      "output_config": {"effort": "low"},
      "messages": [
        {"role": "user", "content": "Classify this ticket as billing, technical, or account: I cannot reset my password."}
      ]
    }'
  ```

  ```python Python theme={null}
  import anthropic

  client = anthropic.Anthropic(
      api_key="YOUR_API_KEY",
      base_url="https://www.anyfast.ai",
  )

  message = client.messages.create(
      model="claude-haiku-5-5",
      max_tokens=1024,
      output_config={"effort": "low"},
      messages=[{"role": "user", "content": "Extract the due date from: Pay by Friday."}],
  )

  for block in message.content:
      if block.type == "text":
          print(block.text)
  ```

  ```python Streaming theme={null}
  import anthropic

  client = anthropic.Anthropic(
      api_key="YOUR_API_KEY",
      base_url="https://www.anyfast.ai",
  )

  with client.messages.stream(
      model="claude-haiku-5-5",
      max_tokens=1024,
      messages=[{"role": "user", "content": "Summarize this support ticket."}],
  ) as stream:
      for text in stream.text_stream:
          print(text, end="")
  ```
</CodeGroup>

## Parameters

| Parameter | Type | Required | Description |
| - | - | - | - |
| `model` | string | Yes | Must be `claude-haiku-5-5`. |
| `messages` | array | Yes | User and assistant turns with text, image, or tool-result content. |
| `max_tokens` | integer | Yes | Maximum output tokens across thinking and visible text; up to 128,000. |
| `output_config` | object | No | Set `effort` to `low`, `medium`, `high`, `xhigh`, or `max`; default `medium`. |
| `thinking` | object | No | Adaptive by default. To disable thinking, omit `output_config.effort` and use `{"type":"disabled"}`. |
| `tools` | array | No | Tool definitions. |
| `tool_choice` | object | No | Use `auto`, `none`, `any`, or a named `tool`. |
| `stream` | boolean | No | Enable SSE streaming; default `false`. |

<Note>
  On AnyFast, `thinking: {"type":"disabled"}` succeeded when `output_config.effort` was omitted; combining it with explicit `effort: "low"` returned `400`. Forced `tool_choice: {"type":"any"}` succeeded and produced a `tool_use` block. Omit custom `temperature`, `top_p`, and `top_k` values, and read response content blocks by `type`.
</Note>

<Note>
  In a tested AnyFast text response, the request ID `claude-haiku-5-5` appeared as `model: "claude-haiku-5.5"`. The response also included `request_id`, `billing.billed_cost_usd` as a string, and `usage.service_tier`. Keep using the hyphen-form ID in requests.
</Note>

<Card title="API Reference" icon="code" href="/api-reference/model-api/anthropic/claude-haiku-5-5">
  Review the Claude Haiku 5.5 request and response fields.
</Card>

<script src="/public/feedback.js" />


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.