Back to Models
qwen logo
qwen/qwen3-max
Not Available

Qwen3 Max

Qwen3-Max is an updated release built on the Qwen3 series, offering major improvements in reasoning, instruction following, multilingual support, and long-tail knowledge coverage compared to the January 2025 version. It delivers higher accuracy in math, coding, logic, and science tasks, follows complex instructions in Chinese and English more reliably, reduces hallucinations, and produces higher-quality responses for open-ended Q&A, writing, and conversation. The model supports over 100 languages with stronger translation and commonsense reasoning, and is optimized for retrieval-augmented generation (RAG) and tool calling, though it does not include a dedicated “thinking” mode.

9/23/2025
262,144 tokens
#79 Text (Korean)

Specifications

Modalities

Input
text
Output
text

Supported Parameters

max_tokens
presence_penalty
response_format
seed
temperature
tool_choice
tools
top_p

Max Output Tokens

32,768

Frequently asked questions

Is Qwen3 Max available on FastMetal?
Not at the moment. Qwen3.8 Max (0902), from the same lab, is available on the FastMetal API today.
What is the context window of Qwen3 Max?
262,144 tokens, shared between the prompt and the response.
How does Qwen3 Max rank?
#85 on the public arena's Japanese board (ELO 1,388). Ranks move as the leaderboard is updated.

Leaderboard

Text
OverallELO: 1,424
#111
JapaneseELO: 1,388
#85
ChineseELO: 1,446
#124
KoreanELO: 1,384
#79
EnglishELO: 1,431
#114
frenchELO: 1,453
#86
germanELO: 1,437
#79
spanishELO: 1,430
#85
russianELO: 1,426
#98
CodingELO: 1,474
#105
MathELO: 1,425
#102
Creative WritingELO: 1,393
#109
Instruction FollowingELO: 1,414
#113
Hard PromptsELO: 1,447
#104
Multi-TurnELO: 1,443
#91