all the models — AI benchmark observatory
← Models

Qwen3 30B A3B

Alibaba Cloud / Qwen Team · open weight · qwen3-30b-a3b

Qwen3-30B-A3B is a smaller Mixture-of-Experts (MoE) model from the Qwen3 series by Alibaba, with 30.5 billion total parameters and 3.3 billion activated parameters. Features hybrid thinking/non-thinking modes, support for 119 languages, and enhanced agent capabilities. It aims to outperform previous models like QwQ-32B while using significantly fewer activated parameters.

Arena HardAIME 2024AIME 2025BFCLGPQALiveCodeBenc

Benchmark scores

BenchmarkScore
Arena Hard0.91
AIME 20240.80
AIME 20250.71
BFCL0.69
GPQA0.66
LiveCodeBench0.63

Pricing

  • Novita$0.10 / $0.44
  • DeepInfra$0.12 / $0.50
  • Fireworks$0.89 / $0.89

Input / output per 1M tokens

AA metrics

No Artificial Analysis link yet.

Arena Elo

  • No Arena snapshot linked.