Human Frontier
63.1
Human-calibrated frontier signal, backed by Arena-style preference evidence and separate from raw AgMoBench benchmark composite scores.
Blended Price
Free/M
Input Price
Free/M
Output Price
Free/M
Speed
—
TTFT
—
Axes
Filters
Providers
vcontinuous-leaderboard · tasks-created:2025-08-01:2026-01-01 · Grok Code Fast 1 · harness swe-rebench-react@tools · 5 attempts · pass@1 · swe-rebench:standardized:128k-or-model-limit · provider Grok Code Fast 1__tools · record Grok Code Fast 1__tools:1754006400000:1767225600000
{
"passAt5": 49.797570850202426,
"developer": "xai",
"instanceType": "model",
"contextPolicy": "128k unless the model supports less",
"attemptsPerTask": 5,
"totalTokenUsage": 1135314.1593588367,
"modelReleaseDate": "2025-08-26",
"scoreStandardError": 0.733757356886186,
"cachedTokenPercentage": null,
"modelReleaseTimestamp": 1756166400000,
"standardContextTokens": 128000,
"averageInstanceCostUsd": 0.060500309698884557,
"potentialContamination": true,
"aggregateTaskRangeTimestamp": {
"to": 1767225600000,
"from": 1754006400000
},
"publishedTaskRangeTimestamp": {
"to": 1767225600000,
"from": 1754006400000
}
}Evidence observed 2026-01-01