kimi-k3
kimi-k3 (Moonshot AI) scored 84.1% composite across 87 tasks. This run covers Australian accounting, Animation, Auth screens, Dashboards, Australian law, Marketing pages, Programming, SVG and graphics, svg-scene, UI components, Websites. Published 2026-07-17.
API single-shotdefault reasoning
Composite
84.1%
95% CI 77.4% - 90.5%
Est. cost
$4.7446
USD
Tokens
349,013
generation
Wall clock
228 min
end to end
Gate pass
-
validated
Separation
-
good vs bad
By domain
Composite score per domain, weakest first. Judge is the vision model’s read, shown for the visual domains.
Marketing page
1 task
31.3%
judge
31.3%
composite
Websites
15 tasks
51.6%
composite
Animation
2 tasks
68.8%
judge
68.8%
composite
Auth screen
1 task
73.4%
judge
73.4%
composite
svg-scene
1 task
75.0%
judge
75.0%
composite
Australian law
5 tasks
80.0%
composite
Dashboard
1 task
84.4%
judge
84.4%
composite
SVG and graphics
13 tasks
86.6%
composite
Australian accounting
23 tasks
92.3%
composite
UI components
19 tasks
99.7%
composite
Programming
6 tasks
100.0%
composite
Outputs
The actual rendered output. Open any tile to view it in a popup, or compare the same task across every model.
53 results
Objective tasks
Programming, Australian legal and accounting, graded by execution. 29 of 34 scored a perfect 100.0%; the rest are below. Open the answer in a popup, or compare it across every model.