Pemikiran adaptif

Pemikiran adaptif adalah cara yang direkomendasikan untuk menggunakan extended thinking dengan Claude Opus 4.6. Alih-alih menetapkan anggaran token pemikiran secara manual, pemikiran adaptif memungkinkan Claude secara dinamis memutuskan kapan dan berapa banyak untuk berpikir berdasarkan kompleksitas setiap permintaan.

Model yang didukung

Pemikiran adaptif didukung pada model berikut:

• Claude Opus 4.6 (claude-opus-4-6)

Cara kerja pemikiran adaptif

Dalam mode adaptif, pemikiran bersifat opsional untuk model. Claude mengevaluasi kompleksitas setiap permintaan dan memutuskan apakah dan berapa banyak untuk berpikir. Pada tingkat effort default (high), Claude hampir selalu akan berpikir. Pada tingkat effort yang lebih rendah, Claude dapat melewati pemikiran untuk masalah yang lebih sederhana.

Pemikiran adaptif juga secara otomatis mengaktifkan interleaved thinking. Ini berarti Claude dapat berpikir di antara panggilan alat, menjadikannya sangat efektif untuk alur kerja agentic.

Cara menggunakan pemikiran adaptif

Atur thinking.type ke "adaptive" dalam permintaan API Anda:

curl https://api.anthropic.com/v1/messages \
     --header "x-api-key: $ANTHROPIC_API_KEY" \
     --header "anthropic-version: 2023-06-01" \
     --header "content-type: application/json" \
     --data \
'{
    "model": "claude-opus-4-6",
    "max_tokens": 16000,
    "thinking": {
        "type": "adaptive"
    },
    "messages": [
        {
            "role": "user",
            "content": "Explain why the sum of two even numbers is always even."
        }
    ]
}'

Pemikiran adaptif dengan parameter effort

Anda dapat menggabungkan pemikiran adaptif dengan parameter effort untuk memandu berapa banyak Claude berpikir. Tingkat effort bertindak sebagai panduan lembut untuk alokasi pemikiran Claude:

import anthropic

client = anthropic.Anthropic()

response = client.messages.create(
    model="claude-opus-4-6",
    max_tokens=16000,
    thinking={
        "type": "adaptive"
    },
    output_config={
        "effort": "medium"
    },
    messages=[{
        "role": "user",
        "content": "What is the capital of France?"
    }]
)

print(response.content[0].text)

Streaming dengan pemikiran adaptif

Pemikiran adaptif bekerja dengan mulus dengan streaming. Blok pemikiran di-stream melalui acara thinking_delta seperti mode pemikiran manual:

import anthropic

client = anthropic.Anthropic()

with client.messages.stream(
    model="claude-opus-4-6",
    max_tokens=16000,
    thinking={"type": "adaptive"},
    messages=[{"role": "user", "content": "What is the greatest common divisor of 1071 and 462?"}],
) as stream:
    for event in stream:
        if event.type == "content_block_start":
            print(f"\nStarting {event.content_block.type} block...")
        elif event.type == "content_block_delta":
            if event.delta.type == "thinking_delta":
                print(event.delta.thinking, end="", flush=True)
            elif event.delta.type == "text_delta":
                print(event.delta.text, end="", flush=True)

Pemikiran adaptif vs manual vs disabled

Pertimbangan penting

Perubahan validasi

Ketika menggunakan pemikiran adaptif, giliran asisten sebelumnya tidak perlu dimulai dengan blok pemikiran. Ini lebih fleksibel daripada mode manual, di mana API memberlakukan bahwa giliran yang diaktifkan pemikiran dimulai dengan blok pemikiran.

Prompt caching

Permintaan berturut-turut menggunakan pemikiran adaptive mempertahankan titik henti prompt cache. Namun, beralih antara mode pemikiran adaptive dan enabled/disabled memecah titik henti cache untuk pesan. Prompt sistem dan definisi alat tetap di-cache terlepas dari perubahan mode.

Tuning perilaku pemikiran

Perilaku pemicu pemikiran adaptif dapat dipromptkan. Jika Claude berpikir lebih atau kurang sering daripada yang Anda inginkan, Anda dapat menambahkan panduan ke prompt sistem Anda:

Extended thinking adds latency and should only be used when it
will meaningfully improve answer quality — typically for problems
that require multi-step reasoning. When in doubt, respond directly.

Kontrol biaya

Gunakan max_tokens sebagai batas keras pada total output (pemikiran + teks respons). Parameter effort memberikan panduan lembut tambahan tentang berapa banyak pemikiran yang Claude alokasikan. Bersama-sama, ini memberi Anda kontrol efektif atas biaya.

Pada tingkat effort high dan max, Claude dapat berpikir lebih ekstensif dan lebih mungkin untuk menghabiskan anggaran max_tokens. Jika Anda mengamati stop_reason: "max_tokens" dalam respons, pertimbangkan untuk meningkatkan max_tokens untuk memberikan lebih banyak ruang kepada model, atau menurunkan tingkat effort.

Bekerja dengan blok pemikiran

Konsep berikut berlaku untuk semua model yang mendukung extended thinking, terlepas dari apakah Anda menggunakan mode adaptif atau manual.

Pemikiran ringkasan

With extended thinking enabled, the Messages API for Claude 4 models returns a summary of Claude's full thinking process. Summarized thinking provides the full intelligence benefits of extended thinking, while preventing misuse.

Here are some important considerations for summarized thinking:

• You're charged for the full thinking tokens generated by the original request, not the summary tokens.
• The billed output token count will not match the count of tokens you see in the response.
• The first few lines of thinking output are more verbose, providing detailed reasoning that's particularly helpful for prompt engineering purposes.
• As Anthropic seeks to improve the extended thinking feature, summarization behavior is subject to change.
• Summarization preserves the key ideas of Claude's thinking process with minimal added latency, enabling a streamable user experience and easy migration from Claude Sonnet 3.7 to Claude 4 and later models.
• Summarization is processed by a different model than the one you target in your requests. The thinking model does not see the summarized output.

Enkripsi pemikiran

Full thinking content is encrypted and returned in the signature field. This field is used to verify that thinking blocks were generated by Claude when passed back to the API.

Here are some important considerations on thinking encryption:

• When streaming responses, the signature is added via a signature_delta inside a content_block_delta event just before the content_block_stop event.
• signature values are significantly longer in Claude 4 models than in previous models.
• The signature field is an opaque field and should not be interpreted or parsed - it exists solely for verification purposes.
• signature values are compatible across platforms (Claude APIs, Amazon Bedrock, and Vertex AI). Values generated on one platform will be compatible with another.

Redaksi pemikiran

Occasionally Claude's internal reasoning will be flagged by our safety systems. When this occurs, we encrypt some or all of the thinking block and return it to you as a redacted_thinking block. redacted_thinking blocks are decrypted when passed back to the API, allowing Claude to continue its response without losing context.

When building customer-facing applications that use extended thinking:

• Be aware that redacted thinking blocks contain encrypted content that isn't human-readable
• Consider providing a simple explanation like: "Some of Claude's internal reasoning has been automatically encrypted for safety reasons. This doesn't affect the quality of responses."
• If showing thinking blocks to users, you can filter out redacted blocks while preserving normal thinking blocks
• Be transparent that using extended thinking features may occasionally result in some reasoning being encrypted
• Implement appropriate error handling to gracefully manage redacted thinking without breaking your UI

Here's an example showing both normal and redacted thinking blocks:

{
  "content": [
    {
      "type": "thinking",
      "thinking": "Let me analyze this step by step...",
      "signature": "WaUjzkypQ2mUEVM36O2TxuC06KN8xyfbJwyem2dw3URve/op91XWHOEBLLqIOMfFG/UvLEczmEsUjavL...."
    },
    {
      "type": "redacted_thinking",
      "data": "EmwKAhgBEgy3va3pzix/LafPsn4aDFIT2Xlxh0L5L8rLVyIwxtE3rAFBa8cr3qpPkNRj2YfWXGmKDxH4mPnZ5sQ7vB9URj2pLmN3kF8/dW5hR7xJ0aP1oLs9yTcMnKVf2wRpEGjH9XZaBt4UvDcPrQ..."
    },
    {
      "type": "text",
      "text": "Based on my analysis..."
    }
  ]
}

When passing thinking and redacted_thinking blocks back to the API in a multi-turn conversation, you must include the complete unmodified block back to the API for the last assistant turn. This is critical for maintaining the model's reasoning flow. We suggest always passing back all thinking blocks to the API. For more details, see the Preserving thinking blocks section.

Harga

For complete pricing information including base rates, cache writes, cache hits, and output tokens, see the pricing page.

The thinking process incurs charges for:

• Tokens used during thinking (output tokens)
• Thinking blocks from the last assistant turn included in subsequent requests (input tokens)
• Standard text output tokens

When using summarized thinking:

• Input tokens: Tokens in your original request (excludes thinking tokens from previous turns)
• Output tokens (billed): The original thinking tokens that Claude generated internally
• Output tokens (visible): The summarized thinking tokens you see in the response
• No charge: Tokens used to generate the summary

Topik tambahan

Halaman extended thinking mencakup beberapa topik secara lebih detail dengan contoh kode khusus mode:

• Penggunaan alat dengan pemikiran: Aturan yang sama berlaku untuk pemikiran adaptif — pertahankan blok pemikiran di antara panggilan alat dan waspadai keterbatasan tool_choice ketika pemikiran aktif.
• Prompt caching: Dengan pemikiran adaptif, permintaan berturut-turut menggunakan mode pemikiran yang sama mempertahankan titik henti cache. Beralih antara mode adaptive dan enabled/disabled memecah titik henti cache untuk pesan (prompt sistem dan definisi alat tetap di-cache).
• Jendela konteks: Bagaimana token pemikiran berinteraksi dengan max_tokens dan batas jendela konteks.

Langkah berikutnya