Claude 3.7 Sonnet vs GPT-4.1: which model should you choose?
We compared Claude 3.7 Sonnet and GPT-4.1 on price, context window and 12 benchmarks. GPT-4.1 has the stronger average benchmark score; GPT-4.1 is cheaper on a blended 3:1 basis.
Claude 3.7 Sonnet
Claude 3.7 Sonnet is a multimodal model from Anthropic. It offers a 200,000-token context window, costs $14.91 per million input tokens and $74.27 per million output tokens, and averages 81.6 across the index's benchmark suite.
- Input
- $14.91 / 1M
- Output
- $74.27 / 1M
- Context
- 200,000
- Avg score
- 81.6
- Open weights
- No
GPT-4.1
GPT-4.1 is a multimodal model from OpenAI. It offers a 1,000,000-token context window, costs $15.71 per million input tokens and $64.52 per million output tokens, and averages 82.4 across the index's benchmark suite.
- Input
- $15.71 / 1M
- Output
- $64.52 / 1M
- Context
- 1,000,000
- Avg score
- 82.4
- Open weights
- No
Specification comparison
| Spec | Claude 3.7 Sonnet | GPT-4.1 | Edge |
|---|---|---|---|
| Input price / 1M | $14.91 | $15.71 | A |
| Output price / 1M | $74.27 | $64.52 | B |
| Context window | 200,000 | 1,000,000 | B |
| Max output | 128,000 | 16,384 | A |
| Average score | 81.6 | 82.4 | B |
| Arena Elo | 1451.9 | 1432.4 | A |
| Open weights | No | No | Tie |
Benchmark comparison
Claude 3.7 Sonnet wins 4, GPT-4.1 wins 8, and 0 benchmarks are tied. The table shows the modelled score for each model and the gap.
| Benchmark | Claude 3.7 Sonnet | GPT-4.1 | Gap | Winner |
|---|---|---|---|---|
| MMLU-Pro | 83.5 | 86.2 | 2.7 | GPT-4.1 |
| GPQA Diamond | 80.2 | 76.1 | 4.1 | Claude 3.7 Sonnet |
| SWE-bench Verified | 70.1 | 72.2 | 2.1 | GPT-4.1 |
| HumanEval | 85.7 | 88.3 | 2.6 | GPT-4.1 |
| MATH-500 | 88.9 | 86.8 | 2.1 | Claude 3.7 Sonnet |
| AIME 2025 | 81.2 | 79.9 | 1.3 | Claude 3.7 Sonnet |
| LiveCodeBench | 75.0 | 77.8 | 2.8 | GPT-4.1 |
| IFEval | 88.5 | 90.4 | 1.9 | GPT-4.1 |
| TruthfulQA | 81.7 | 81.8 | 0.1 | GPT-4.1 |
| HellaSwag | 87.6 | 90.3 | 2.7 | GPT-4.1 |
| MMMU | 75.5 | 76.3 | 0.8 | GPT-4.1 |
| Arena Elo | 1451.9 | 1432.4 | 19.5 | Claude 3.7 Sonnet |
Price analysis
On a blended 3:1 input-to-output basis, Claude 3.7 Sonnet costs about $29.75 per million tokens and GPT-4.1 costs about $27.91. GPT-4.1 is roughly 6% cheaper than Claude 3.7 Sonnet. At high volume, that gap compounds quickly, so price should carry real weight unless the quality difference is decisive.
GPT-4.1 has the larger context window at 1,000,000 tokens, compared with 200,000 for Claude 3.7 Sonnet. That matters for long documents, large codebases and multi-step agent traces.
Verdict
Across 12 shared benchmarks, Claude 3.7 Sonnet wins 4, GPT-4.1 wins 8, and 0 are tied. On the index average, GPT-4.1 leads Claude 3.7 Sonnet by 0.8 points (82.4 versus 81.6).
The price picture favours GPT-4.1. The context picture favours GPT-4.1. Those two dimensions often matter more than a few benchmark points once a model is good enough for the task.
If you need a single default, start with GPT-4.1 for quality-sensitive work, and check whether its price holds at your expected volume before committing.
Choose Claude 3.7 Sonnet if
- You need the larger 128,000-token output limit.
Choose GPT-4.1 if
- You want the higher average benchmark score in this index.
- Unit cost matters and you want the cheaper blended price.
- You need the larger 1,000,000-token context window.
- The workload needs multi-step reasoning.
FAQ
Is Claude 3.7 Sonnet better than GPT-4.1?
On the index average, GPT-4.1 leads Claude 3.7 Sonnet by 0.8 points (82.4 versus 81.6). The better choice depends on price, context and your task. Claude 3.7 Sonnet wins 4 benchmarks, GPT-4.1 wins 8.
Which is cheaper, Claude 3.7 Sonnet or GPT-4.1?
GPT-4.1 is cheaper on a blended 3:1 basis at about $27.91 per million tokens, versus $29.75 for Claude 3.7 Sonnet.
Which has the larger context window?
GPT-4.1 has the larger context window at 1,000,000 tokens. Claude 3.7 Sonnet has 200,000 tokens.
Are these models open weights?
Claude 3.7 Sonnet is proprietary; GPT-4.1 is proprietary.
Which should I choose for production?
If you need a single default, start with GPT-4.1 for quality-sensitive work, and check whether its price holds at your expected volume before committing.