GPT-4.1 vs GPT-5.4
Head to head on 13 shared benchmarks, with price, context window, and release dates.
Benchmark scores
64evals · sort by any column, ⤢ to expand fullscreen
| Model | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GPT-5.4 OpenAI | 1.47 | 65.3% | 1836elo | 47.6% | 87.2% | 74.8% | 11.3% | 48.4% | 68.3% | 47.1% | 74.0% | 37.9% | 36.0% | - | - | - | - | - | - | 13.1% | - | - | 81.0% | 34.1% | 9.0% | - | 1272elo | 78.3 | 78.5% | 92063 | 82.0% | 77.5% | 79.3% | 70.2% | 82.6% | 94.1% | 88.1% | - | - | - | - | - | 82.8% | - | - | 41.3% | - | 77.5% | - | - | 86 | 52.2% | 56.0% | 43.3% | - | -0.61 | 40.8 | - | - | - | 80 | 67.4% | - | - | 60.3% |
| GPT-4.1 OpenAI | 1.55 | 63.1% | 1417elo | 5.5% | 60.9% | 66.6% | 4.2% | 43.0% | 65.9% | 38.1% | 75.1% | 13.6% | 47.1% | 0.0% | 52.4% | 43.7% | 34.7% | 26.7% | 61.5% | - | 1.64 | 83.7% | - | - | - | 12.4% | - | - | - | - | - | - | - | - | - | - | - | 45.7% | 70.8% | 0.49 | 0.49 | 91.3% | - | 32.1% | 32.1% | - | 69.4% | - | 80.6% | 50.5% | - | - | - | - | 8.57 | - | - | 31.1% | 39.6% | 10.0% | - | - | 92.0% | 92.0% | 48.0% |
2 / 2 models
| Attribute | GPT-4.1 OpenAI | GPT-5.4 OpenAI |
|---|---|---|
| Description | GPT-4.1 is an AI model from OpenAI. | GPT-5.4 is an AI model from OpenAI. |
| Family | GPT | GPT |
| Params | - | - |
| Open weights | closed | closed |
| License | proprietary | proprietary |
| Released | 14 Apr 2025 | 5 Mar 2026 |
| Context | 1,047,576 | 1,050,000 |
| Input $/1M | $2.00 | $2.50 |
| Output $/1M | $8.00 | $15.00 |
| Cache read $/1M | $0.500 | $0.250 |
| Cache write $/1M | - | - |
| Speed | 0 tok/s | 0 tok/s |
| Latency | 0.00 s | 0.00 s |
| Modalities | image, text, file | text, image, file |
| Providers | - | - |
| Publisher | OpenAI | OpenAI |
| Reported scores | ScholarSearch8.57 Social Sciences & Humanities (%) GPQA Diamond66.6% (AA) Humanity's Last Exam (HLE)4.2% (AA) IFBench43.0% (AA) | GPQA Diamond74.8% (AA) SciCode47.1% (AA) τ²-bench (Tau²-bench)36.0% (AA) Terminal-Bench (Hard)37.9% (AA) Vibe Code Bench v1.167.4% |
FAQ
- Is GPT-4.1 or GPT-5.4 better?
- Across the 13 benchmarks both models report on Sophon, GPT-5.4 leads on 10 of them. Which one is "better" depends on the benchmark - the table above shows every shared score.
- How do GPT-4.1 and GPT-5.4 compare on GPQA Diamond?
- GPT-4.1 scores 66.6% and GPT-5.4 scores 74.8% on GPQA Diamond.
- Which is cheaper, GPT-4.1 or GPT-5.4?
- GPT-4.1 is cheaper at $8.00 per million output tokens against $15.00 for GPT-5.4 - about 1.9x.
- When were GPT-4.1 and GPT-5.4 released?
- GPT-4.1 was released 14 Apr 2025 by OpenAI. GPT-5.4 was released 5 Mar 2026 by OpenAI.
Comparing something else? Build your own side-by-side.