kimi · 5 arms · 5 effortskimi-k3

One model across every effort and arm. Each cell is a run — open it for the full card. Failure chips mark runs with no valid slot.

effort strip

arm \ effort	none	low	high	max	disabled
api	—	open run ↗ 26K tokens8 runs · all valid	open run ↗ 60.2K tokens8 runs · all valid	—	open run ↗ 31.3K tokens8 runs · all valid
cursor	—	open run ↗ 13.9K tokens4 runs · all valid	open run ↗ 48.7K tokens4 runs · all valid	open run ↗ 65.7K tokens4 runs · all valid	—
cursor-xl	—	open run ↗ 11.6K tokens3 of 4 valid	open run ↗ 47.4K tokens4 runs · all valid	no valid slot no valid slot0 of 4 valid	—
go	open run ↗ 74.8K tokens4 runs · all valid	—	—	—	—
opencode	open run ↗ — tokens4 runs · all valid	—	—	—	—

Output completion tokens per effort, per arm. A read on how much each model spends as effort rises — not a quality axis.

4.7×low → max output

APIcursorcursor-xlGo (native)opencode