AI Hub
← All models

Llama 3.2 90B Instruct vs Qwen3.6 Plus

Meta vs Alibaba — benchmarks, pricing, and capabilities side by side.

  • Qwen3.6 Plus has the higher intelligence index (88.2 vs 54)
  • Qwen3.6 Plus is cheaper ($0.33 vs $0.35 per 1M input)
  • Llama 3.2 90B Instruct is faster
  • Qwen3.6 Plus has a larger context window (1M)
Llama 3.2 90B InstructQwen3.6 Plus
Intelligence index5488.2
DeveloperMetaAlibaba
TypeMultimodalMultimodal
AccessOpen weightsAPI only
Context window128,000 tokens1,000,000 tokens
Input price$0.35 / 1M$0.33 / 1M
Output price$0.40 / 1M$1.95 / 1M
Speed100 tok/s52 tok/s
ReleasedSeptember 25, 2024April 2, 2026
Parameters90000000000
Input modalitiesText, Image, Video
Output modalitiesText

Shared benchmarks

Llama 3.2 90B Instruct
Qwen3.6 Plus
GPQA Diamond
46.7
88.2
Humanity’s Last Exam
4.9
25.7