DeepSeek · LIVEBENCH 2026-06-25

DeepSeek V4.1 Flash Max Effort

Published benchmark results for this specific model configuration.

deepseek-v4.1-flash-max

Overall score81.1 / 100

Cost / successful task$0.029USD · benchmark workload

Weight accessOpen weights

Where this configuration performs

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

Reasoning86.7

Coding80.0

Agentic Coding77.3

Mathematics93.3

Data Analysis79.3

Language81.2

Instruction Following70.0

Task-level results

theory of mind80.8zebra puzzle100.0spatial98.0logic with navigation68.0code generation77.5code completion82.6javascript81.8typescript60.0python90.0AMPS Hard98.0integrals with game90.0math comp96.1olympiad89.1consecutive events88.7tablejoin49.0tablereformat100.0connections100.0plot unscrambling61.6typos82.0paraphrase70.7simplify67.1story generation72.7summarize69.7

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