in the loop

Qwen3-235B-A22B
#123 of 274 scored.

AlibabaApr 28, 2025Open weights

Capabilities Index

Range 135.5–140.8

139.3

Context

Tokens

—

Input

Per 1M tokens

—

Output

Per 1M tokens

—

Benchmarks

How it scores.

  • GPQA Diamond60.9%

    100th of 186

Everything else Epoch has for it

  • Lech Mazur Writing83.0%
  • MATH level 568.9%
  • Fiction.LiveBench67.7%
  • Aider polyglot59.6%
  • DTBench59.6%
  • WeirdML37.3%
  • LMCA29.4%
  • SimpleBench17.2%

On the index

Around it.

  1. #120Qwen3-30B-A3B-Thinking (Jul 2025)Alibaba139.6
  2. #121Qwen3.5-9BAlibaba · $0.10 in139.5
  3. #122GPT-5 nanoOpenAI · $0.05 in139.4
  4. #123Qwen3-235B-A22BAlibaba139.3
  5. #124DeepSeek-R1DeepSeek · $0.70 in139.0
  6. #125Qwen3-235B-A22B-Instruct (Jul 2025)Alibaba138.9
  7. #126Qwen3-32BAlibaba · $0.08 in138.5

Sources