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

# Extended Thinking Anthropic

> Konfigurasikan anggaran token penalaran dan tangkap blok thinking pada Claude.

Extended Thinking pada endpoint `/anthropic/v1/messages` memungkinkan model [Claude](https://www.anthropic.com/claude?utm_source=neosantara-docs\&utm_medium=referral) memikirkan masalah kompleks sebelum memberikan jawaban. Anda dapat menentukan batas anggaran token pemikiran menggunakan objek `thinking`.

<CodeGroup>
  ```python Python (Anthropic SDK) icon="python" theme={"theme":{"light":"ayu-dark","dark":"catppuccin-latte"}}
  from anthropic import Anthropic
  import os

  client = Anthropic(
      base_url="https://api.neosantara.xyz/anthropic",
      api_key=os.environ["NEOSANTARA_API_KEY"]
  )

  response = client.messages.create(
      model="claude-sonnet-4-6",
      max_tokens=4096,
      thinking={
          "type": "enabled",
          "budget_tokens": 2048
      },
      messages=[
          {"role": "user", "content": "Berapa banyak huruf 'r' dalam kata 'strawberry'?"}
      ]
  )

  for block in response.content:
      if block.type == "thinking":
          print(f"[Proses Pemikiran]:\n{block.thinking}\n")
      elif block.type == "text":
          print(f"[Jawaban Akhir]:\n{block.text}")
  ```

  ```bash cURL icon="terminal" theme={"theme":{"light":"ayu-dark","dark":"catppuccin-latte"}}
  curl -X POST https://api.neosantara.xyz/anthropic/v1/messages \
    -H "x-api-key: $NEOSANTARA_API_KEY" \
    -H "anthropic-version: 2023-06-01" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "claude-sonnet-4-6",
      "max_tokens": 4096,
      "thinking": {
        "type": "enabled",
        "budget_tokens": 2048
      },
      "messages": [
        {"role": "user", "content": "Berapa banyak huruf r dalam kata strawberry?"}
      ]
    }'
  ```
</CodeGroup>

## Aturan Parameter `thinking`

* **Nilai `max_tokens` Wajib Lebih Besar:** Nilai `max_tokens` harus selalu lebih besar dari nilai `budget_tokens` agar tersedia ruang token untuk jawaban akhir.
* **Batas Minimum Anggaran:** Nilai minimum untuk `budget_tokens` adalah 1024 token.
* **Struktur Blok Respons:** Respons terdiri dari urutan blok konten di mana blok `type: "thinking"` muncul sebelum blok `type: "text"`.

## Langkah Selanjutnya

| Kebutuhan                     | Panduan                                                         |
| :---------------------------- | :-------------------------------------------------------------- |
| Tool Use Anthropic            | [Anthropic Tool Use](/id/gateway/anthropic-messages/tool-use)   |
| Streaming Anthropic           | [Streaming Anthropic](/id/gateway/anthropic-messages/streaming) |
| Penalaran di Chat Completions | [Penalaran OpenAI](/id/gateway/chat-completions/reasoning)      |


## Related topics

- [Anthropic Messages API](/id/gateway/anthropic-messages.md)
- [Penalaran (Reasoning)](/id/gateway/chat-completions/reasoning.md)
- [Penalaran Responses API](/id/gateway/responses-api/reasoning.md)