Phi-4-Final: benchmark results

Provider: Microsoft.

Aggregate Rating 1489 ± 42, rank #2522 of 10916 rated models, from 11 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
LLMEval-Fair - Literature5.53Discipline Score (0-10)44.8
LLMEval-Fair - Management6.3Discipline Score (0-10)43.1
LLMEval-Fair - Science6.27Discipline Score (0-10)39.7
LLMEval-Fair - Engineering5.8Discipline Score (0-10)38.8
LLMEval-Fair - Medicine5.43Discipline Score (0-10)37.9
LLMEval-Fair - Overall59.93Absolute Score (0-100)32.8
LLMEval-Fair - Economics6.47Discipline Score (0-10)31
LLMEval-Fair - History5.5Discipline Score (0-10)27.6
LLMEval-Fair - Law6.53Discipline Score (0-10)26.7
LLMEval-Fair - Military5.87Discipline Score (0-10)25.9
LLMEval-Fair - Education6.23Discipline Score (0-10)17.2

Interactive version: theaggregate.ai/model?slug=phi-4-final · How It Works · Data refreshed daily, snapshot 2026-11-08.