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 Instruct | Qwen3.6 Plus | |
|---|---|---|
| Intelligence index | 54 | 88.2 |
| Developer | Meta | Alibaba |
| Type | Multimodal | Multimodal |
| Access | Open weights | API only |
| Context window | 128,000 tokens | 1,000,000 tokens |
| Input price | $0.35 / 1M | $0.33 / 1M |
| Output price | $0.40 / 1M | $1.95 / 1M |
| Speed | 100 tok/s | 52 tok/s |
| Released | September 25, 2024 | April 2, 2026 |
| Parameters | 90000000000 | — |
| Input modalities | — | Text, Image, Video |
| Output modalities | — | Text |
Shared benchmarks
Llama 3.2 90B Instruct
Qwen3.6 Plus
GPQA Diamond
46.7
88.2
Humanity’s Last Exam
4.9
25.7