Output tokens per task
#27 fewest of 7374k
13.9
#59 of 81 models
Bars show where K2 Horizon 32B sits among the 81 ranked models on each benchmark (right is best); the tick is the field median.
Across 17 results on every current benchmark that measured K2 Horizon 32B, not only this ranking’s. The percentage is its average standing among the 81 models in this ranking that took the same benchmarks: 100% beat them all, 0% trailed them all. A benchmark in two categories counts in both; a category needs two benchmarks.
74k
Not measured for K2 Horizon 32B: cost per task, time per task, output speed, time to first token (no published API price yet).
Cost per task is what a subscriber pays: the model’s plan price, a share of API price. Models no plan sells are at API price.
Agents & computer use
Leader: Gemini 4 Neon · 79.3%
Professional work
Leader: Sonnet 5.5 · 1815
Science & research
Leader: Fable 5.2 · 76.5%
Coding
Leader: Fable 5.2 · 49.1%
Professional work
Leader: Gemini 4 Neon · 33.4%
Knowledge & accuracy
Leader: Fable 5.2 · 80.2%
Knowledge & accuracy
Leader: Fable 5.2 · 79.1%
Knowledge & accuracy
Leader: Gemini 4 Carbon · 95.8%
Science & research
Leader: GPT-6.2 Astra · 78.6%
Agents & computer use
Leader: Kimi K3.1 · 67.2%