Beta version: *Information might not be fully accurate. Please report any discrepancies.

AlibabaVerified4 benchmarks

Qwen 2.5 Max

Released 2025-01-29Unknown Architecture

Training: 2024-09

Latest Data

Unknown

Context Window

131k

tokens

Input Cost

$1.60

per 1M tokens

Output Cost

$6.40

per 1M tokens

Parameters

Unknown

model footprint

Benchmark Provenance

Performance Analysis // Verified Benchmarks

87.9*/ 100

Verified

Massive Multitask Language Understanding covers 57 subjects across STEM, the humanities, social sciences, and more.

68.5*/ 100

Verified

Challenging competition mathematics problems (AIME/IMO level).

73.2*/ 100

Verified

Functional correctness of synthesized programs from docstrings.

69*/ 100

Verified

A more robust and harder version of MMLU, focusing on complex reasoning and STEM subjects.