OpenAI
GPT-4.1 benchmark results
14 results on 10 benchmarks from 6 sources.
Also listed as: gpt-4.1-2025-04-14 · gpt-4.1-20250414 · GPT 4.1 · GPT 4.1 (2025-04-14) · GPT-4.1-2025-04-14 (FC) · GPT-4.1-2025-04-14 (Prompt)
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1415 | 140 | Overall, style control on; 50314 votes; 95% CI 1411–1418“gpt-4.1-2025-04-14” | Official leaderboard | Oct 8, 2026 |
| LMArena Vision | 1214 | 78 | Overall, style control on; 41020 votes; 95% CI 1207–1221“gpt-4.1-2025-04-14” | Official leaderboard | Oct 2, 2026 |
| GPQA Diamond | 66.9% | 169 | Epoch AI run, accuracy | Independent run | Apr 14, 2025 |
| MMLU-Pro | 81.8% | 55 | Overall accuracy; data source: Self-Reported | Vendor-reported | Mar 11, 2026 |
| SWE-bench Verified | 48.5% | 32 | Epoch AI run (Epoch scaffold) | Independent run | Feb 8, 2026 |
| SWE-bench Verified | 39.6% | 84 | agent: mini-SWE-agent“gpt-4.1-20250414” | Official leaderboard | Jul 26, 2025 |
| AIME (OTIS Mock AIME 2024–2025) | 38.3% | 182 | Epoch AI run, OTIS Mock AIME 2024–2025 | Independent run | Apr 14, 2025 |
| MATH Level 5 | 83.0% | 33 | Epoch AI run, accuracy | Independent run | Apr 14, 2025 |
| Humanity's Last Exam | 5.4% | 45 | Compiled by Epoch AI | Third-party compilation via Epoch AI | Apr 14, 2025 |
| τ²-bench | 56.0% | 19 | airline domain, pass^1; submitted by Sierra; v0.1.3 | Official leaderboard | Jun 9, 2025 |
| τ²-bench | 74.0% | 18 | retail domain, pass^1; submitted by Sierra; v0.1.3 | Official leaderboard | Jun 9, 2025 |
| τ²-bench | 34.0% | 21 | telecom domain, pass^1; submitted by Sierra; v0.1.3 | Official leaderboard | Jun 9, 2025 |
| BFCL v4 (function calling) | 54.0% | 20 | Overall accuracy; native function calling (FC)“GPT-4.1-2025-04-14 (FC)” | Official leaderboard | Oct 17, 2026 |
| BFCL v4 (function calling) | 39.4% | 45 | Overall accuracy; prompt-based“GPT-4.1-2025-04-14 (Prompt)” | Official leaderboard | Oct 17, 2026 |
FAQ
Where do these GPT-4.1 scores come from?
From Epoch AI Benchmarking Hub, Berkeley Function Calling Leaderboard (BFCL), LMArena leaderboard dataset, MMLU-Pro leaderboard (TIGER-Lab), SWE-bench leaderboards (swebench.com), and τ²-bench leaderboard submissions. Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.