Model Comparison
OpenAI's GPT-4.1 costs less per intelligence point, even though Anthropic's Claude 4.5 Sonnet (Non-reasoning) scores higher.
Data last updated March 4, 2026
GPT-4.1 delivers more intelligence per dollar, while Claude 4.5 Sonnet (Non-reasoning) leads on raw benchmark scores. Claude 4.5 Sonnet (Non-reasoning) costs $0.03 per request vs $0.018 for GPT-4.1 (at 5K input / 1K output tokens). GPT-4.1 scores proportionally higher on mathematical reasoning (AIME: 0.44), while Claude 4.5 Sonnet (Non-reasoning)'s scores skew toward general knowledge (MMLU-Pro: 0.86). The question is whether Claude 4.5 Sonnet (Non-reasoning)'s higher scores justify paying 67% more.
| Metric | Claude 4.5 Sonnet (Non-reasoning) | GPT-4.1 |
|---|---|---|
| Intelligence IndexComposite score from MMLU-Pro, GPQA, and AIME. Higher is better. | 37.1 | 26.3 |
| MMLU-ProGeneral knowledge and reasoning. Higher is better. | 0.9 | 0.8 |
| GPQAGraduate-level science questions. Higher is better. | 0.7 | 0.7 |
| Output tokens/secTokens generated per second. Higher means faster responses. | 45.3 | 77.2 |
| Time to first tokenSeconds until first token. Lower is better. | 1.20s | 0.51s |
| Context windowMax tokens per request. Larger handles more text. | 1,000,000 | 1,047,576 |
List prices as published by the provider. Not adjusted for token efficiency.
| Metric | Claude 4.5 Sonnet (Non-reasoning) | GPT-4.1 |
|---|---|---|
| Input price / 1M tokens | $3.00 | $2.00 |
| Output price / 1M tokens | $15.00 | $8.00 |
| Cache hit price / 1M tokens | $0.30 | $0.50 |
Cost per IQ point based on a typical request of 5,000 input and 1,000 output tokens.
Cheaper (list price)
GPT-4.1
Higher Benchmarks
Claude 4.5 Sonnet (Non-reasoning)
Better Value ($/IQ point)
GPT-4.1
Claude 4.5 Sonnet (Non-reasoning)
$0.0008 / IQ point
GPT-4.1
$0.0007 / IQ point
GPT-4.1 is 67% cheaper per request than Claude 4.5 Sonnet (Non-reasoning). GPT-4.1 is cheaper on both input ($2.0/M vs $3.0/M) and output ($8.0/M vs $15.0/M). The 67% price gap matters at scale but is less significant for low-volume use cases. This comparison assumes a typical request of 5,000 input and 1,000 output tokens (5:1 ratio). Actual ratios vary by workload — chat and completion tasks typically run 2:1, code review around 3:1, document analysis and summarization 10:1 to 50:1, and embedding workloads are pure input with no output tokens.
Claude 4.5 Sonnet (Non-reasoning) scores higher overall (37.1 vs 26.3). Claude 4.5 Sonnet (Non-reasoning) leads on MMLU-Pro (0.86 vs 0.81), GPQA (0.73 vs 0.67). GPT-4.1 scores proportionally higher on AIME (mathematical reasoning) relative to its MMLU-Pro, while Claude 4.5 Sonnet (Non-reasoning)'s scores are more weighted toward general knowledge. Claude 4.5 Sonnet (Non-reasoning)'s GPQA score of 0.73 makes it stronger for technical and scientific tasks.
GPT-4.1 is 70% faster at 77.2 tokens per second compared to Claude 4.5 Sonnet (Non-reasoning) at 45.3 tokens per second. GPT-4.1 also starts generating sooner at 0.51s vs 1.20s time to first token. The speed difference matters for chatbots but is less relevant in batch processing.
GPT-4.1 has a 5% larger context window at 1,047,576 tokens vs Claude 4.5 Sonnet (Non-reasoning) at 1,000,000 tokens. That's roughly 1,396 vs 1,333 pages of text. The extra context capacity in GPT-4.1 matters for document analysis and long conversations.
GPT-4.1 offers 18% better value at $0.0007 per intelligence point compared to Claude 4.5 Sonnet (Non-reasoning) at $0.0008. GPT-4.1 is cheaper, which offsets Claude 4.5 Sonnet (Non-reasoning)'s higher benchmark scores to deliver more value per dollar. If raw benchmark scores matter less than cost for your use case, GPT-4.1 is the efficient choice.
With prompt caching, GPT-4.1 is 57% cheaper per request than Claude 4.5 Sonnet (Non-reasoning). Caching saves 45% on Claude 4.5 Sonnet (Non-reasoning) and 42% on GPT-4.1 compared to standard input prices. Both models benefit from caching at similar rates, so the uncached price comparison holds.
Pricing verified against official vendor documentation. Updated daily. See our methodology.
Related Comparisons
Create an account, install the SDK, and see your first margin data in minutes.
See My Margin DataNo credit card required