AI KNOWLEDGE DESK

Models · entities · concepts · comparisons · practical tools

GETLLMS.ORG

Claude Opus 5.5

Claude Opus 5.5 is Anthropic's current default recommendation for long-running agentic coding and knowledge work.

Language ModelCoding AssistantAgent WorkflowsAdaptive ThinkingLong ContextAnthropic
Claude API and documented Claude product surfaces
$4/1M input, $20/1M output, $0.20/1M cache reads; 5m cache writes $5 and 1h cache writes $8
Commercial

🚀Function Overview

Claude Opus 5.5 lowers Opus 5 token and cache-read prices while retaining a 1M context window, 128K output, vision, tools, and adaptive thinking.

Key Features

  • Official Claude API model ID claude-opus-5-5
  • 1M-token context window and 128K maximum output
  • Text and image input with text output and tool use
  • Adaptive thinking is always on; medium is the documented default effort
  • $4 input and $20 output per million tokens
  • $0.20 cache reads, $5 five-minute cache writes, and $8 one-hour cache writes per million tokens
  • Anthropic positions Fable 5.1 above it for the hardest reasoning and long-horizon agent work
  • Current model overview recommends Opus 5.5 as the starting point for most workloads

Use Cases

  • Supervised feature work across several files
  • Long-running coding agents
  • Knowledge work with document or image input
  • Tool-using debugging and code review
  • Migration from Claude Opus 5

⚙️Input Parameters

model

string

Use claude-opus-5-5.

messages

array

Anthropic Messages API conversation payload.

effort

string

Choose low, medium, high, xhigh, or supported session max; medium is the documented default.

tools

array

Optional tool definitions for the active Claude surface.

💡Usage Examples

Example 1

Input Parameters

{
  "model": "claude-opus-5-5",
  "messages": [
    {
      "role": "user",
      "content": "Trace this bug, implement the smallest fix, and run the relevant checks."
    }
  ],
  "effort": "high"
}

Output Results

Claude Opus 5.5 returns content blocks and tool calls. Compare validated task cost, retries, output tokens, and cache-hit behavior on your own workflow.

Quick Actions

Technical Specifications

Hardware Type
Claude API and documented Claude product surfaces
Commercial Use
Supported
Pricing
$4/1M input, $20/1M output, $0.20/1M cache reads; 5m cache writes $5 and 1h cache writes $8