Qwen2.5 Max vs Sonar Reasoning Pro
Alibaba vs Perplexity — benchmarks, pricing, and capabilities side by side.
- •Sonar Reasoning Pro has the higher intelligence index (95.7 vs 63.6)
- •Qwen2.5 Max is cheaper ($1.60 vs $2.00 per 1M input)
| Qwen2.5 Max | Sonar Reasoning Pro | |
|---|---|---|
| Intelligence index | 63.6 | 95.7 |
| Developer | Alibaba | Perplexity |
| Type | LLM | Multimodal |
| Access | — | API only |
| Context window | — | 128,000 tokens |
| Input price | $1.60 / 1M | $2.00 / 1M |
| Output price | $6.40 / 1M | $8.00 / 1M |
| Speed | 50 tok/s | — |
| Released | January 28, 2025 | March 7, 2025 |
| Parameters | — | — |
| Input modalities | — | Text, Image |
| Output modalities | — | Text |
Shared benchmarks
Qwen2.5 Max
Sonar Reasoning Pro
MATH-500
83.5
95.7