Description

Technical specs

Training cut-off date

Input formats

Supported language

Supported Azure regions

Sample JSON response

Input

{
  "model": "claude-opus-4-8",
  "max_tokens": 4096,
  "messages": [
    {
      "role": "user",
      "content": "Analyze this code for security vulnerabilities and suggest fixes."
    }
  ]
}

Output

{
  "content": [
    {
      "text": "I'll analyze this code for security vulnerabilities...",
      "type": "text"
    }
  ],
  "id": "msg_example",
  "model": "claude-opus-4-8",
  "role": "assistant",
  "stop_reason": "end_turn",
  "type": "message",
  "usage": {
    "input_tokens": 42,
    "output_tokens": 256
  }
}

Model architecture

Long context

Optimizing model performance

Additional assets

• Claude Documentation : Visit Claude documentation for a wealth of resources on model capabilities, prompting techniques, use case guidelines, and more.
• Adaptive Thinking Guide : Understand how best to use adaptive thinking with Claude.
• Claude Prompting Resources : Check out Anthropic's prompting tools and guides to learn how to craft prompts that elicit more helpful, nuanced responses.
• Claude Cookbooks : Check out example code for a variety of complex tasks, such as RAG from various web sources, making SQL queries, function calling, multimodal prompting, and more.

Key capabilities

About this model

Key model capabilities

• Adaptive thinking is an upgrade to extended thinking that gives Claude the freedom to think as much or as little as needed depending on the task and effort level. We've also introduced a new effort level, xhigh.
• 1M context window: Claude Opus 4.8 supports a 1M token context window, enabling reasoning over entire codebases and carrying multi-day project context forward without frequent compaction.
• High-resolution image input: Claude Opus 4.8 processes images at up to 2576px / 3.75MP delivering sharper results on screenshots, documents, and computer use.

Pricing

Use cases

Key use cases

• Long-running agents: Claude Opus 4.8 sets a new bar for agentic workflows, pairing improved tool use with creative problem-solving across multi-step tasks. It works independently for longer, unblocking itself across multi-stage projects and complex dependency chains with minimal oversight.
• Coding: Claude Opus 4.8 is the best coding model for real-world work, and runs independently for longer than previous Opus models. It reads codebases like a subject matter expert and plans before it edits. On long, multi-stage tasks it keeps track of dependencies, gets unstuck on its own, and persists over hours with stronger memory and longer context.
• Enterprise workflows: Claude Opus 4.8 is built for scalable workflows. It is better at following instructions, staying in scope, and producing professional-grade outputs. It reasons across long documents, checks its work, and manages multi-stage projects end-to-end with professional polish on spreadsheets, slides, and docs.
• Financial analysis: Claude Opus 4.8 brings deeper reasoning and precision to financial workflows like investment research, earnings analysis, reading dense filings and charts at high fidelity and carrying context across an entire reporting cycle.
• Cybersecurity: Claude Opus 4.8 brings deeper reasoning to security workflows like threat intelligence synthesis, vulnerability finding, alert triage, and incident response. It holds long traces and large codebases in context to help catch subtle patterns and complex attack vectors.
• Computer use: Claude Opus 4.8 is our most capable generally available computer-use model, pairing improved vision with deep reasoning for multi-step tasks that span multiple applications and require planning and judgment.

Out of scope use cases

Distribution channels

• Claude API: For developers interested in building agents, Claude Opus 4.8 is available on the Claude Platform.
• Claude Code: Use Claude Opus 4.8 with Anthropic's industry-leading coding agent, Claude Code.

More information