phi 4

Microsoft · microsoft/phi-4

ProviderMicrosoft
Familyphi
Typellm-code
Statusactive
Released2024-12-11

Reported benchmark scores

These scores carry one collection date for the whole card (2026-04) and a source list rather than a source per score, so they cannot be attributed individually. They are shown as reported and are not verified evidence. Benchmark pages carry the reviewed, dated evidence where it exists.
BenchmarkCatalogue standingScore
AlpacaEvalunassessed32.5
Arena Elo — Codingunassessed1250.0
Arena Elo — Hard Promptsunassessed1277.2
Arena Elo — Mathunassessed1260.0
Arena Elo — Overall (Text)unassessed1240.0
Arena Elo — Style Controlunassessed1255.4
Artificial Analysis Intelligence Indexunassessed68.0
Artificial Analysis Output Speedunassessed88.0
BIG-Bench Hardunassessed54.3
GPQA Diamondunassessed52.5
GSM8Kunassessed93.5
HumanEvalunassessed78.5
IFEvalunassessed78.0
LiveCodeBenchunassessed23.1
MATH-500unassessed82.5
MedQAunassessed77.8
MGSM (Multilingual Grade School Math)unassessed80.6
MMLU: Astronomyunassessed68.8
MMLU: Biology (subcategory)unassessed77.8
MMLU: Business Ethicsunassessed70.2
MMLU: Chemistry (subcategory)unassessed70.5
MMLU: Clinical Knowledgeunassessed76.2
MMLU: Computer Science (subcategory)unassessed78.5
MMLU: Jurisprudenceunassessed66.8
MMLU: Physics (unresolved key)unassessed74.2
MMLU-Prounassessed68.5
MMLU: Professional Accountingunassessed61.2
MMLU: Professional Lawunassessed65.5
MT-Benchunassessed8.2
MultiPL-Eunassessed72.0
MultiPL-E: C++unassessed72.1
MultiPL-E: C#unassessed72.8
MultiPL-E: Gounassessed67.5
MultiPL-E: Javaunassessed76.5
MultiPL-E: JavaScriptunassessed74.2
MultiPL-E: Juliaunassessed52.5
MultiPL-E: Kotlin (unconfirmed)unassessed62.8
MultiPL-E: Luaunassessed45.5
MultiPL-E: Perlunassessed40.8
MultiPL-E: PHPunassessed69.5
MultiPL-E: Pythonunassessed79.2
MultiPL-E: Runassessed48.2
MultiPL-E: Rubyunassessed55.2
MultiPL-E: Rustunassessed60.8
MultiPL-E: Scalaunassessed50.5
MultiPL-E: Swiftunassessed55.8
MultiPL-E: TypeScriptunassessed73.8
MuSRunassessed23.5
WildBenchunassessed58.2

Data

This card as JSON · Edit on GitHub