GPT-6.2 Sol
Everyday generalist. Everyday coding, general work and tool-using agents.
Which reasoning effort setting?
controls how long GPT-6.2 Sol thinks before answering. Token prices never change; higher settings spend more and raise task cost.
Start at Xhigh. It scores 55.7 against 57.3 at the top setting, for 36% less per benchmark task. Rule: cheapest setting within 2 points of the model’s best score.
At Xhigh, GPT-6.2 Sol thinks for about 11.1K tokens per task, then writes about 13.3K more as its answer: $0.65 per task, 3.5× Low, for +15.1 points.
Output rate: $10.00/1M at every effort level. Thinking tokens are billed as output.
| Effort | Index | Per task | Thinking | All output | × Low | Time |
|---|---|---|---|---|---|---|
| Low | 40.6 | $0.18 | 1.3K | 6K | 1.00× | 1.6 min |
| Medium | 52.7 | $0.41 | 4.4K | 13.7K | 2.21× | 3.6 min |
| High | 54.9 | $0.56 | 8.1K | 20.2K | 3.07× | 5.2 min |
| XhighStart here | 55.7 | $0.65 | 11.1K | 24.3K | 3.52× | 6.1 min |
| Max | 57.3 | $1.01 | 22.8K | 40.4K | 5.51× | 9.8 min |
Per task of the Artificial Analysis Intelligence Index: the average over its nine evaluations, counting input, cache, and output tokens at list prices. These are long, agent-style benchmark tasks, much longer than a chat message.
Where GPT-6.2 Sol sits on the frontier
Intelligence Index against one 100K-input, 10K-output request.