Inference
EN
Alibaba

Qwen2.5-Max

Qwen2.5-Max is a standard model from Alibaba. It has 6 scored results in 3 categories. It has no index.

  • Closed weights
  • Standard
  • Input: text
  • Released 2025-01
UnrankedEnough results for an index.
IndexAll categories, out of 100.UnrankedMedian 51.0
CoverageIndex weight that has results.45%
SpeedOutput tokens per second.—Median 67/s
First tokenSeconds to the first output token.—Median 1.3 s
Input / 1MUS dollars per 1M input tokens.—Median $0.60
Output / 1MUS dollars per 1M output tokens.—Median $2.55
ContextMaximum tokens in one request.—Median 262K
EloLMArena rating and rank.1367±4 #193Median 1422

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.None counted——
CodingCode writing and repair.1 counted21.5—
ReasoningLogic problems and puzzles.None counted——
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.2 counted35.4—
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.3 counted26.9—

Results

6 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
MATH level 5Math67.2% ±1.135 / 88
GPQA diamond · Epoch AIKnowledge56.1% ±2.8132 / 217
Massive Multitask Language Understanding ProfessionalKnowledge76.1%46 / 75
OTIS Mock AIME 2024-2025Math16.1% ±4.1149 / 200
Graduate-Level Google-Proof Q&AKnowledge60.1%109 / 141
FrontierMath-2025-02-28-PrivateMath1.0% ±0.659 / 69
LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for CodeCoding38.7%11 / 12
FrontierMath-2025-02-28-PublicMath0.0%33 / 64

Sources

BenchLMCC BY-NC 4.0Epoch AICC BY

Questions

Is Qwen2.5-Max open source?
No. Its weights are not published.
When was Qwen2.5-Max released?
January 2025.

More from Alibaba

All models
Qwen2.5-1MUnrankedQwen2.5-VL-32BUnrankedQwen PlusUnrankedQwen2.5 VL 72B InstructUnrankedQwQ-32B36.3