Kimi K3.1 Pro has landed a score of 55 on the Artificial Analysis Intelligence Index, a result that places the model on the Pareto line in the firm's latest independent analysis. The benchmark, published by Artificial Analysis, ranks the model fourth out of 215 models in intelligence, well above the median score of 27 for comparable models.
The result matters because the Pareto line marks the models that offer the best trade-off between intelligence and cost. According to the analysis, Kimi K3.1 Pro delivers leading intelligence at a moderate price, putting it in the same competitive group as frontier models from OpenAI, Anthropic, Google and DeepSeek.
Kimi K3.1 Pro specs and benchmark scores
Artificial Analysis reports that Kimi K3.1 Pro scores 55 on the Intelligence Index v4.3.2, which incorporates 10 evaluations including Humanity's Last Exam, SciCode, Terminal-Bench 4.0 and GDPval-AA v2.1. The score places it fourth overall and well above the median of 27 for its class.
The analysis describes the model as very verbose. When evaluated on the Intelligence Index, it generated 150 million tokens, compared with a median of 87 million. Pricing is listed at $2.00 per million input tokens and $10.00 per million output tokens, both described as moderate relative to comparable models, with a 90% cache discount available.
Speed is the weaker point. At 73.5 output tokens per second, Kimi K3.1 Pro is slower than the average of 82 tokens per second. The evaluation cost an average of $2.11 per task on the Intelligence Index.
- Intelligence Index score: 55 (median: 27), ranked 4th of 215 models
- Output speed: 73.5 tokens per second (average: 82)
