Alibaba · LIVEBENCH 2026-06-25

Qwen 3.8 Max

Published benchmark results for this specific model configuration.

qwen3.8-max

Overall score78.5 / 100

Cost / successful task$0.275USD · benchmark workload

Weight accessOpen weights

Where this configuration performs

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

Reasoning88.2

Coding72.9

Agentic Coding64.6

Mathematics91.3

Data Analysis78.4

Language79.7

Instruction Following74.1

Task-level results

theory of mind78.8zebra puzzle100.0spatial100.0logic with navigation74.0code generation71.8code completion73.9javascript77.3typescript56.7python60.0AMPS Hard98.0integrals with game81.0math comp95.1olympiad91.2consecutive events87.1tablejoin50.1tablereformat98.0connections94.5plot unscrambling58.6typos86.0paraphrase73.3simplify67.3story generation76.5summarize79.2

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