Qwen2.5 Max

Qwen2.5 Max

Alibaba · Released Jan 28, 2025 Efficient
Intelligence #284 / 583
31.1 Our Score
Speed #232 / 276
48.3 tokens / sec
Input #482 / 583
$1.60 per 1M tokens
Output #485 / 583
$6.40 per 1M tokens
Context
— Not reported

Analysis Summary

Qwen2.5 Max is Alibaba's flagship from the Qwen 2.5 generation, released January 2025. With an intelligence index of 10.2, MMLU-Pro of 0.762, and GPQA of 0.587, it performs well in the mid-upper tier of the 2025 landscape. Its LiveCodeBench score of 0.359 and SciCode of 0.337 indicate solid coding and scientific reasoning capability.

For business use, it suits complex content generation, structured analysis, and moderate coding assistance. At $1.60 input and $6.40 output per million tokens, pricing is reasonable for its capability tier. However, enterprise adoption outside Alibaba's primary markets may face friction around support, compliance, and integration compared to providers with broader global infrastructure.

Teams with no provider constraints and a need for strong general reasoning at a mid-range price will find it capable. Those prioritising enterprise support and ecosystem maturity may prefer alternatives from providers with more established global deployment.

Assessed June 17, 2026

Editorial notes

Qwen2.5 Max from Alibaba delivers strong GPQA and MMLU-Pro scores with a competitive LiveCodeBench result, but a -4 point regional accessibility adjustment applies given its provider context.

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?

Performance Profile

Intelligence2.4Technical0Value6Content5.5
Intelligence 2.4/10
Technical 0/10
Content 5.5/10
Value 6/10

How Qwen2.5 Max compares

Qwen2.5 Max ranks #217 of 380 AI models we track for overall intelligence. At $1.60 per million input tokens it is cheaper than 17% of comparable models.

Performance Indices

Source: Artificial Analysis

10.2 Intelligence Index

Benchmark Scores

Intelligence

GPQA Diamond 58.7% Graduate-level scientific reasoning
HLE 4.5% Humanity's Last Exam
MMLU Pro 76.2% Multi-task language understanding
MATH 500 83.5% Mathematical problem-solving
AIME 23.3% Competition mathematics
SciCode 33.7% Scientific computing

Technical

LiveCodeBench 35.9% Live coding evaluation

Benchmark data from Artificial Analysis and Hugging Face

How does Qwen2.5 Max stack up?

Compare side-by-side with other efficient models.

Compare Models

Model Information

ProviderAlibaba
Release Date January 28, 2025
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $1.60 $0.001600
Output $6.40 $0.006400

Leaderboard Categories

Frequently asked questions about Qwen2.5 Max

How much does Qwen2.5 Max cost?

Qwen2.5 Max costs $1.60 per million input tokens and $6.40 per million output tokens.

Who created Qwen2.5 Max?

Qwen2.5 Max is developed by Alibaba and was released on January 28, 2025.