- Intelligence
- 22.0
- Rank 218 of 744 measured models
- Coding
- 21.0
- Rank 222 of 298 measured models
- Output speed
- 159 t/s
- Rank 73 of 403 measured models
- Price/1M
- $1.08
- Rank 387 of 659 measured models
- Context window
- 256K
- Rank 518 of 1199 measured models
Kimi K2 Thinking scores 22.0 on the Artificial Analysis Intelligence Index, ranking 218th of 744 measured models. The closest models on this index are Gemini 3.5 Flash-Lite (22.2), GLM-4.7 (22.2), and o3-pro (21.9). At $1.08 per million tokens (3:1 input/output blended price), it is cheaper than o3-pro ($35), more expensive than Gemini 3.5 Flash-Lite ($0.85) and GLM-4.7 ($1). With an output speed of 159 t/s, it is faster than GLM-4.7 (37.9 t/s), slower than Gemini 3.5 Flash-Lite (435 t/s).
Additional benchmarks1
Design Arena · Website48.8%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Gemini 3.5 Flash-Lite
Google · $0.85 · 435 t/s
22.2+0.2$0.85435 t/sCompare - GLM-4.7
Z.AI · $1 · 37.9 t/s
22.2+0.2$137.9 t/sCompare - o3-pro
OpenAI · $35 · —
21.9−0.1$35—Compare - G9v3-39A5B
AI9Stars · — · —
21.8−0.2——Compare
