Model Comparison
OpenAI's GPT-4o costs less per intelligence point, even though Anthropic's Claude 4 Opus (Non-reasoning) scores higher.
Data last updated March 4, 2026
GPT-4o delivers more intelligence per dollar, while Claude 4 Opus (Non-reasoning) leads on raw benchmark scores. Claude 4 Opus (Non-reasoning) costs $0.15 per request vs $0.0225 for GPT-4o (at 5K input / 1K output tokens). Claude 4 Opus (Non-reasoning) scores proportionally higher on mathematical reasoning (AIME: 0.56), while GPT-4o's scores skew toward general knowledge (MMLU-Pro: 0.75). The question is whether Claude 4 Opus (Non-reasoning)'s higher scores justify the 7x price premium.
| Metric | Claude 4 Opus (Non-reasoning) | GPT-4o |
|---|---|---|
| Intelligence IndexComposite score from MMLU-Pro, GPQA, and AIME. Higher is better. | 22.2 | 17.3 |
| MMLU-ProGeneral knowledge and reasoning. Higher is better. | 0.9 | 0.8 |
| GPQAGraduate-level science questions. Higher is better. | 0.7 | 0.5 |
| AIMEMathematical problem solving. Higher is better. | 0.6 | 0.2 |
| Output tokens/secTokens generated per second. Higher means faster responses. | 36.2 | 116.8 |
| Time to first tokenSeconds until first token. Lower is better. | 1.34s | 0.44s |
| Context windowMax tokens per request. Larger handles more text. | 1,000,000 | 128,000 |
List prices as published by the provider. Not adjusted for token efficiency.
| Metric | Claude 4 Opus (Non-reasoning) | GPT-4o |
|---|---|---|
| Input price / 1M tokens | $15.00 | $2.50 |
| Output price / 1M tokens | $75.00 | $10.00 |
| Cache hit price / 1M tokens | $1.50 | $1.25 |
Cost per IQ point based on a typical request of 5,000 input and 1,000 output tokens.
Cheaper (list price)
GPT-4o
Higher Benchmarks
Claude 4 Opus (Non-reasoning)
Better Value ($/IQ point)
GPT-4o
Claude 4 Opus (Non-reasoning)
$0.0068 / IQ point
GPT-4o
$0.0013 / IQ point
GPT-4o is dramatically cheaper — 7x less per request than Claude 4 Opus (Non-reasoning). GPT-4o is cheaper on both input ($2.5/M vs $15.0/M) and output ($10.0/M vs $75.0/M). At a fraction of the cost, GPT-4o saves significantly in production workloads. 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 Opus (Non-reasoning) scores higher overall (22.2 vs 17.3). Claude 4 Opus (Non-reasoning) leads on MMLU-Pro (0.86 vs 0.75), GPQA (0.7 vs 0.54), AIME (0.56 vs 0.15). Claude 4 Opus (Non-reasoning) scores proportionally higher on AIME (mathematical reasoning) relative to its MMLU-Pro, while GPT-4o's scores are more weighted toward general knowledge. If mathematical reasoning matters, Claude 4 Opus (Non-reasoning)'s AIME score of 0.56 gives it an edge.
GPT-4o is significantly faster — 116.8 tokens per second vs Claude 4 Opus (Non-reasoning) at 36.2 tokens per second. GPT-4o also starts generating sooner at 0.44s vs 1.34s time to first token. For interactive use cases, GPT-4o's speed advantage translates to noticeably lower latency.
Claude 4 Opus (Non-reasoning) has a much larger context window — 1,000,000 tokens vs GPT-4o at 128,000 tokens. That's roughly 1,333 vs 170 pages of text. Claude 4 Opus (Non-reasoning)'s window can handle entire codebases or book-length documents; GPT-4o works better for shorter inputs.
GPT-4o offers dramatically better value — $0.0013 per intelligence point vs Claude 4 Opus (Non-reasoning) at $0.0068. GPT-4o is cheaper, which offsets Claude 4 Opus (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-4o is the efficient choice.
With prompt caching, GPT-4o is dramatically cheaper — 5x less per request than Claude 4 Opus (Non-reasoning). Caching saves 45% on Claude 4 Opus (Non-reasoning) and 28% on GPT-4o compared to standard input prices. Claude 4 Opus (Non-reasoning) benefits more from caching. If your workload has repetitive prompts, Claude 4 Opus (Non-reasoning)'s cache discount gives it a bigger cost advantage than list prices suggest.
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