xAI · LIVEBENCH 2026-06-25

Grok Build 0.1

Published benchmark results for this specific model configuration.

grok-build-0.1

Overall score67.8 / 100

Cost / successful task$0.024USD · benchmark workload

Weight accessNot reported

Where this configuration performs

Category averages from the same release, on a 0–100 scale.

Reasoning76.4

Coding65.4

Agentic Coding45.8

Mathematics78.4

Data Analysis70.8

Language72.5

Instruction Following65.2

Task-level results

theory of mind69.2zebra puzzle68.3spatial98.0logic with navigation70.0code generation63.4code completion67.4javascript59.1typescript33.3python45.0AMPS Hard66.0integrals with game66.0math comp95.1olympiad86.6consecutive events69.7tablejoin42.7tablereformat100.0connections84.0plot unscrambling53.4typos80.0paraphrase58.4simplify62.6story generation72.5summarize67.4

Source files: Scores ↗ Categories ↗ Costs ↗ . Missing cost is unknown, never free. Compare configurations within the same release.