Human Frontier
86.0
Human-calibrated frontier signal, backed by Arena-style preference evidence and separate from raw AgMoBench benchmark composite scores.
Blended Price
$1.20/M
Input Price
$0.60/M
Output Price
$3.00/M
Speed
—
TTFT
—
Axes
Filters
Providers
vcontinuous-leaderboard · tasks-created:2026-01-01:2026-03-01 · Kimi K2.5 · harness swe-rebench-react@tools · 5 attempts · pass@1 · swe-rebench:standardized:128k-or-model-limit · provider Kimi K2.5__tools · record Kimi K2.5__tools:1767225600000:1772323200000
{
"passAt5": 60.952380952380956,
"developer": "moonshot",
"instanceType": "model",
"contextPolicy": "128k unless the model supports less",
"attemptsPerTask": 5,
"totalTokenUsage": 1361950.016043956,
"modelReleaseDate": "2026-01-26",
"scoreStandardError": 0.7979212617639928,
"cachedTokenPercentage": 92.2213727004562,
"modelReleaseTimestamp": 1769385600000,
"standardContextTokens": 128000,
"averageInstanceCostUsd": 0.19543760866117216,
"potentialContamination": true,
"aggregateTaskRangeTimestamp": {
"to": 1772323200000,
"from": 1767225600000
},
"publishedTaskRangeTimestamp": {
"to": 1772323200000,
"from": 1767225600000
}
}Evidence observed 2026-03-01