Claude 4.1Released 2025-09-10
Claude Opus 4.1
Enhanced Opus 4 model with improved instruction following
Family
Claude 4.1
Release Date
2025-09-10
Best For
Complex reasoning, instruction-heavy tasks, professional workflows
Overview
Claude Opus 4.1 is an enhanced version of Opus 4 with improved instruction following and task execution. At $15/$75 per MTok, it targets professional workflows that demand the highest quality output and can justify the cost. Suitable for complex reasoning, expert analysis, and tasks requiring precise adherence to detailed instructions.
Pricing
Input / 1M tokens
$15.00
Output / 1M tokens
$75.00
Context Window
200K
Max Output
32K
API model ID:
claude-opus-4-1Key Features
✓Improved instruction following over Opus 4
✓200K token context window
✓32K token max output
✓Enhanced task execution and planning
✓Strong coding and analytical capabilities
Use Cases
Complex professional workflows
Detailed instruction-heavy tasks
Expert-level analysis and reporting
Research requiring precise output formatting
Benchmarks
| Benchmark | Score |
|---|---|
| Graduate-Level Reasoning (GPQA) | 81.5% |
| Coding (SWE-bench Verified) | 46.0% |
| MMLU (5-shot) | 90.5% |