Institute of Foundation Models
K2 Horizon 375B A23B
- Intelligence
- 17.8
- Rank 286 of 771 measured models
- Coding
- 61.5
- Rank 80 of 299 measured models
- Output speed
- 96.1 t/s
- Rank 197 of 426 measured models
- Context window
- 524K
- Rank 365 of 1301 measured models
- Cost per task
- $0
- Rank 1 of 219 measured models
K2 Horizon 375B A23B scores 17.8 on the Artificial Analysis Intelligence Index, ranking 286th of 771 measured models. The closest models on this index are Qwen3.6 35B A3B (18.2), Beam 1.0 (17.7), and Qwen3.5 122B A10B (17.7). With an output speed of 96.1 t/s, it is slower than Qwen3.6 35B A3B (122 t/s), Beam 1.0 (330 t/s), and Qwen3.5 122B A10B (128 t/s).
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Qwen3.6 35B A3B
Alibaba · $0.844 · 122 t/s
18.2+0.4$0.844122 t/sCompare - Beam 1.0
Reflection · $0.60 · 330 t/s
17.7−0.1$0.60330 t/sCompare - Qwen3.5 122B A10B
Alibaba · $1.10 · 128 t/s
17.7−0.1$1.10128 t/sCompare - Kimi K3
Moonshot AI · $6 · 46.4 t/s
17.4−0.4$646.4 t/sCompare
