GPT-6.1 Terra
Visual automation. Low-cost screen work, visual reasoning and drafting.
Which reasoning effort setting?
controls how long GPT-6.1 Terra thinks before answering. Token prices never change; higher settings spend more and raise task cost.
Start at Max. The top setting is also the best value here. Rule: cheapest setting within 2 points of the model’s best score.
At Max, GPT-6.1 Terra thinks for about 39.4K tokens per task, then writes about 20.5K more as its answer: $0.96 per task, 13.5× Low, for +21.9 points.
Output rate: $5.00/1M at every effort level. Thinking tokens are billed as output.
| Effort | Index | Per task | Thinking | All output | × Low | Time |
|---|---|---|---|---|---|---|
| None | 31.0 | $0.23 | — | — | — | 2.7 min |
| Low | 29.4 | $0.071 | 950 | 4.4K | 1.00× | 1.4 min |
| Medium | 38.2 | $0.17 | 4.1K | 10.9K | 2.37× | 3.1 min |
| High | 42.4 | $0.26 | 7.6K | 16.7K | 3.63× | 4.6 min |
| Xhigh | 45.3 | $0.40 | 14.2K | 26.2K | 5.62× | 7.3 min |
| MaxStart here | 51.3 | $0.96 | 39.4K | 59.9K | 13.48× | 16 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-6.1 Terra sits on the frontier
Intelligence Index against one 100K-input, 10K-output request.