Kimi K2.5 vs Sonar Pro
Moonshot AI vs Perplexity — benchmarks, pricing, and capabilities side by side.
- •Kimi K2.5 has the higher intelligence index (87.9 vs 58.8)
- •Kimi K2.5 is cheaper ($0.40 vs $3.00 per 1M input)
- •Kimi K2.5 has a larger context window (262K)
| Kimi K2.5 | Sonar Pro | |
|---|---|---|
| Intelligence index | 87.9 | 58.8 |
| Developer | Moonshot AI | Perplexity |
| Type | Multimodal | Multimodal |
| Access | Open weights | API only |
| Context window | 262,144 tokens | 200,000 tokens |
| Input price | $0.40 / 1M | $3.00 / 1M |
| Output price | $1.90 / 1M | $15.00 / 1M |
| Speed | 35 tok/s | — |
| Released | January 27, 2026 | March 7, 2025 |
| Parameters | 1T (32B active) | — |
| Input modalities | Text, Image | Text, Image |
| Output modalities | Text | Text |
Shared benchmarks
Kimi K2.5
Sonar Pro
GPQA Diamond
87.9
57.8
Humanity’s Last Exam
29.4
7.9