Phi-3.5-MoE-instruct vs Qwen3.6 Max
Microsoft vs Alibaba — benchmarks, pricing, and capabilities side by side.
- •Qwen3.6 Max has the higher intelligence index (88.8 vs 49.8)
| Phi-3.5-MoE-instruct | Qwen3.6 Max | |
|---|---|---|
| Intelligence index | 49.8 | 88.8 |
| Developer | Microsoft | Alibaba |
| Type | LLM | LLM |
| Access | Open weights | API only |
| Context window | — | 262,144 tokens |
| Input price | — | $1.04 / 1M |
| Output price | — | $6.24 / 1M |
| Speed | — | 36 tok/s |
| Released | August 23, 2024 | April 27, 2026 |
| Parameters | 60000000000 | — |
| Input modalities | — | Text |
| Output modalities | — | Text |
Shared benchmarks
Phi-3.5-MoE-instruct
Qwen3.6 Max
GPQA Diamond
36.8
88.8