GPT-6.1 Sol vs GPT-6 Sol
OpenAI has officially announced GPT-6.1 Sol, its answer to Claude Opus 5.5. Here is how it compares to GPT-6 Sol:
Benchmarks
| Benchmark | GPT-6.1 Sol | GPT-6 Sol | Difference |
|---|---|---|---|
| DeepSWE v1.1 | 75.2% | 68.8% | GPT-6.1 Sol +6.4 points |
| GDP.pdf | 32.0% | 28.0% | GPT-6.1 Sol +4.0 points |
| AutomationBench 1.0.6 | 36.1% | 33.2% | GPT-6.1 Sol +2.9 points |
| OSWorld 2.0 Offline | 71.4% | 64.4% | GPT-6.1 Sol +7.0 points |
| Terminal-Bench Science 0.1 | 57.0% | 27.6% | GPT-6.1 Sol +29.4 points |
| Factual error rate (lower is better) | 4.1% | 4.5% | GPT-6.1 Sol 0.4 points lower |
Pricing and specs
| Specification | GPT-6.1 Sol | GPT-6 Sol | Difference |
|---|---|---|---|
| Input, per 1M tokens | $2 | $2 | Same |
| Cached input, per 1M tokens | $0.10 | $0.20 | 50% cheaper |
| Output, per 1M tokens | $10 | $10 | Same |
| Context window | 1.05M tokens | 1.05M tokens | Same |
| Max output | 128K tokens | 128K tokens | Same |
Artificial Analysis benchmarks
| Benchmark | GPT-6.1 Sol (Max) | GPT-6 Sol (Max) | Difference |
|---|---|---|---|
| Intelligence Index v4.3.2 | 51.8 | 47.5 | GPT-6.1 Sol +4.3 |
| AA-Briefcase v1.1 | 1564.2 Elo | 1482.8 Elo | GPT-6.1 Sol +81.4 Elo |
| GDPval-AA v2.1 | 1575.1 Elo | 1487 Elo | GPT-6.1 Sol +88.1 Elo |
| AutomationBench-AA | 64.9% | 62% | GPT-6.1 Sol +2.9 points |
| Terminal-Bench 4.0 | 56.1% | 43.9% | GPT-6.1 Sol +12.2 points |
| SciCode | 54.2% | 57.6% | GPT-6 Sol +3.4 points |
| Humanity's Last Exam | 52.9% | 48% | GPT-6.1 Sol +4.9 points |
| GDP.pdf | 31.0% | 24.8% | GPT-6.1 Sol +6.2 points |
| CritPt | 31.7% | About 31% | GPT-6.1 Sol about +0.7 points |
| AA-Omniscience | 41.5 | 27.1 | GPT-6.1 Sol +14.4 |
| AA-LCR v1.1 | 83.0% | 83.7% | GPT-6 Sol +0.7 points |