A

Claude Opus 4

Anthropic
Tools Reasoning Vision
200K
Context
$15.00
Input /1M
$75.00
Output /1M
32K
Max Output
Run API Call Open in Chat

Claude Opus 4

Overview

Property Value
Model ID anthropic/claude-opus-4
Permanent Slug anthropic/claude-4-opus-20250514
Provider Anthropic
Created May 22, 2025
Context Length 200,000 tokens
Max Completion Tokens 32,000

Description

Claude Opus 4 is benchmarked as the world's best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows. It sets new benchmarks in software engineering, achieving leading results on SWE-bench (72.5%) and Terminal-bench (43.2%). Opus 4 supports extended, agentic workflows, handling thousands of task steps continuously for hours without degradation.

For more details, see the Anthropic announcement blog post.

Technical Specifications

Input/Output Modalities

Type Supported Modalities
Context Window 200,000 tokens
Input Text, Images, Files
Output Text

Capabilities

  • Supports Reasoning: Yes
  • Tool Use: Yes (with tool_choice parameter)
  • Multipart Support: Yes
  • Extended Agentic Workflows: Yes - handles thousands of task steps continuously for hours without degradation

Pricing

Amazon Bedrock Provider

Type Price per Token Price per Million Tokens
Input $0.000015 $15.00
Output $0.000075 $75.00
Cached Input (Read) $0.0000015 $1.50
Cached Input (Write) $0.00001875 $18.75
Image Processing $0.024/image -

Supported Parameters

Parameter Description
temperature Controls randomness in outputs
top_p Nucleus sampling parameter
top_k Top-k sampling parameter
stop Stop sequences
max_tokens Maximum tokens to generate
tools Tool definitions for function calling
tool_choice Control tool selection behavior
reasoning Enable reasoning capabilities
include_reasoning Include reasoning in response

API Usage Example

curl https://api.langmart.ai/v1/chat/completions \
  -H "Authorization: Bearer $LANGMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-opus-4",
    "messages": [
      {
        "role": "user",
        "content": "Explain the concept of recursion in programming."
      }
    ],
    "max_tokens": 1024,
    "temperature": 0.7
  }'
Model Context Notes
anthropic/claude-sonnet-4 200K Faster, more cost-effective variant
anthropic/claude-3.5-sonnet 200K Previous generation
anthropic/claude-3-opus 200K Previous flagship model

Providers

Amazon Bedrock (Primary)

Property Value
Provider Model ID us.anthropic.claude-opus-4-20250514-v1:0
Context Length 200,000 tokens
Max Completion Tokens 32,000
Moderation Required
Data Policy No training use, no prompt retention

Default Provider Order

LangMart uses the following provider preference order:

  1. Google Vertex AI
  2. Amazon Bedrock
  3. Anthropic (Direct)

Performance Metrics

Benchmark Scores

Benchmark Score Notes
SWE-bench 72.5% Leading result at release
Terminal-bench 43.2% Leading result at release

Usage Analytics (Sample: December 22, 2025)

Metric Value
Daily Requests 11,784
Prompt Tokens 186.9M
Completion Tokens 5.8M
Multimodal Prompts 1,132 (with images)

Key Differentiators

  1. Best-in-class Coding Performance: Benchmarked as the world's best coding model at release
  2. Extended Agent Workflows: Handles thousands of task steps continuously without degradation
  3. Sustained Performance: Maintains quality over long-running complex tasks
  4. Reasoning Support: Native support for reasoning and chain-of-thought capabilities
  5. Multimodal Input: Supports text, images, and file inputs

Source

Last Updated: December 23, 2025