Abacus.AI · LIVEBENCH 2026-06-25

Smaug Mini

Published benchmark results for this specific model configuration.

smaug-mini

Overall score76.9 / 100

Cost / successful task$0.099USD · benchmark workload

Weight accessOpen weights

Where this configuration performs

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

Reasoning82.8

Coding75.4

Agentic Coding60.8

Mathematics89.7

Data Analysis78.8

Language77.0

Instruction Following73.9

Task-level results

theory of mind69.2zebra puzzle94.0spatial98.0logic with navigation70.0code generation70.4code completion80.4javascript59.1typescript53.3python70.0AMPS Hard98.0integrals with game76.0math comp94.1olympiad90.9consecutive events88.5tablejoin51.9tablereformat96.1connections91.7plot unscrambling55.3typos84.0paraphrase76.2simplify65.4story generation78.0summarize76.1

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