Anthropic Claude 4.7 Opus Documentation

Claude 4.7 Opus is built for sophisticated, asynchronous agent pipelines and persistent knowledge work. It provides elite performance on tasks where execution unfolds over time, offering unmatched coherence across ultra-long outputs and extended multi-step sessions.

Key Capabilities

Asynchronous Agent Optimization: Engineered to reliably execute multi-stage workflows without human intervention or conversational drift.
Complex Project Orchestration: Seamlessly handles end-to-end execution across multi-file codebases and complex infrastructure tasks.
Persistent Knowledge Work: Exceptional at drafting comprehensive documentation, building structured presentations, and executing deep data analyses.
Advanced Multi-Stage Debugging: Traces, isolates, and fixes subtle bugs that span across disparate modules or services.
High-Fidelity Judgment: Exhibits superior follow-through and context retention over extended timelines, making it a reliable default for technical design.

Request Parameters

To interact with this model via an OpenAI-compatible gateway or OpenRouter endpoint:

Parameter	Type	Required	Description
`model`	`string`	Yes	Use `"claude-opus-4-7"` or `"anthropic/claude-4-7-opus"`.
`messages`	`array`	Yes	Array of standard conversation messages (role/content).
`max_tokens`	`integer`	No	Limits the generation length. Supports extended output targets.
`temperature`	`float`	No	Recommended: `0.0` for precise architectural logic; `0.5` for document drafting.

Thinking Mode Configuration

claude-opus-4-7 supports Adaptive Thinking — the model dynamically decides when and how much to think. The legacy "type": "enabled" + budget_tokens parameters are removed on this model (will return a 400 error).

Important Constraints

If thinking is included, temperature, top_p, and top_k are not supported (will return 400).
max_tokens should be set generously (e.g., ≥ 64000) when using "effort": "xhigh" or "max".
Thinking content is omitted by default. Add "display": "summarized" to see the model's reasoning.

Effort Levels

Value	Description
`"low"`	Fast; minimal thinking for simple tasks
`"medium"`	Balanced speed and depth
`"high"`	Deep reasoning — recommended default
`"xhigh"`	Best for coding and agentic tasks
`"max"`	Maximum reasoning for the hardest problems

Request Example

json Copy

{
  "model": "claude-opus-4-7",
  "max_tokens": 64000,
  "thinking": {
    "type": "adaptive",
    "display": "summarized"
  },
  "output_config": {
    "effort": "xhigh"
  },
  "messages": [
    {
      "role": "user",
      "content": "Your question here"
    }
  ]
}

Note: "type": "enabled" and budget_tokens are fully removed on this model and will return a 400 error. Use "type": "adaptive" + output_config.effort instead.

claude-opus-4-7

Input

Output