AI Hub
← All models

Phi-3.5-mini-instruct vs Sonar Reasoning Pro

Microsoft vs Perplexity — benchmarks, pricing, and capabilities side by side.

  • Sonar Reasoning Pro has the higher intelligence index (95.7 vs 46)
  • Phi-3.5-mini-instruct is cheaper ($0.10 vs $2.00 per 1M input)
Phi-3.5-mini-instructSonar Reasoning Pro
Intelligence index4695.7
DeveloperMicrosoftPerplexity
TypeLLMMultimodal
AccessOpen weightsAPI only
Context window128,000 tokens128,000 tokens
Input price$0.10 / 1M$2.00 / 1M
Output price$0.10 / 1M$8.00 / 1M
Speed23 tok/s
ReleasedAugust 23, 2024March 7, 2025
Parameters3800000000
Input modalitiesText, Image
Output modalitiesText