Noometry

Model comparison

GPT-4.5 vs Phi-4

GPT-4.5 is the stronger model overall, scoring 37.2 to 31.2 on the Noometry Index.

Last verified . 29 shared benchmarks.

GPT-4.5 OpenAI

37.2

Rank #208 Confirmed

Phi-4 Microsoft

31.2

Rank #279 Confirmed

Summary

  • They share 29 benchmarks with published results for both. GPT-4.5 scores higher in 7 categories and Phi-4 in 2 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where GPT-4.5 leads 56.9 to 40.5.
  • The biggest single-benchmark swing is LiveBench Coding: 75.2% for GPT-4.5 and 30.7% for Phi-4.
  • Phi-4 has downloadable open weights; the other is API-only.

Side by side

GPT-4.5Phi-4
ProviderOpenAIMicrosoft
Noometry Index37.231.2
Released2025-02-272024-12-11
WeightsProprietaryOpen
Context window—128K
Max output—4K
Input $ / M tokens—$0.07
Output $ / M tokens—$0.14
Results tracked4237

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding GPT-4.5 leads

GPT-4.5: 42.2 (#109), Phi-4: 34.4 (#239)

BenchmarkGPT-4.5Phi-4
LiveBench Coding75.2%30.7%
LMArena Coding13961231
Aider Polyglot44.9%—
WeirdML39.4%—
BigCodeBench Instruct—45.5%
BigCodeBench Complete—55.4%

Agentic & Tool Use GPT-4.5 leads

GPT-4.5: 27.9 (#97), Phi-4: 22.8 (#128)

BenchmarkGPT-4.5Phi-4
Berkeley Function Calling Leaderboard—28.8%
Cybench17.5%—
BALROG—11.6%

Reasoning Phi-4 leads

GPT-4.5: 13.9 (#330), Phi-4: 17.7 (#291)

BenchmarkGPT-4.5Phi-4
LiveBench Reasoning71.1%47.8%
LMArena Hard Prompts14031220
LiveBench Data Analysis64.3%45.2%
Epoch Capabilities Index136.74130.42
LiveBench69%41.6%
ARC-AGI-20.8%—
SimpleBench34.5%—
ARC-AGI-110.3%—
Chess Puzzles—1%
EnigmaEval3.2%—
ForecastBench61.7—

Math GPT-4.5 leads

GPT-4.5: 32.6 (#211), Phi-4: 20.8 (#285)

BenchmarkGPT-4.5Phi-4
OTIS Mock AIME 2024-202537.8%13.8%
LiveBench Math69.3%42%
LMArena Math14121246
MATH Level 578.6%64.9%

Knowledge Too close to call

GPT-4.5: 32.5 (#211), Phi-4: 32.6 (#209)

BenchmarkGPT-4.5Phi-4
GPQA Diamond68.7%56.1%
Confabulations13.6%29.4%
LMArena Expert13941203
Humanity's Last Exam5.4%—
Vectara Hallucination Rate—3.7%
MMLU—84.8%

Multimodal Not comparable

GPT-4.5: 37.6 (#71), Phi-4: —

BenchmarkGPT-4.5Phi-4
LMArena Vision1195—
VPCT45%—

Multilingual GPT-4.5 leads

GPT-4.5: 52.5 (#83), Phi-4: 37.2 (#237)

BenchmarkGPT-4.5Phi-4
LMArena Non-English14131197
LMArena Chinese14211212
LMArena French14181224
LMArena German14571222
LMArena Japanese14161158
LMArena Korean13921151
LMArena Russian14191209
LMArena Spanish—1234

Instruction Following GPT-4.5 leads

GPT-4.5: 72.6 (#134), Phi-4: 60.4 (#251)

BenchmarkGPT-4.5Phi-4
LiveBench Instruction Following72.3%58.4%
LMArena Instruction Following14041201

Long Context GPT-4.5 leads

GPT-4.5: 40.4 (#155), Phi-4: 36.9 (#226)

BenchmarkGPT-4.5Phi-4
LMArena Longer Query14061217
Fiction.LiveBench63.9%—

Writing & Preference GPT-4.5 leads

GPT-4.5: 56.9 (#134), Phi-4: 40.5 (#244)

BenchmarkGPT-4.5Phi-4
LMArena Text14171217
LMArena Creative Writing13941182
Short-Story Creative Writing75.6%62.6%
LMArena Multi-Turn14441206
LiveBench Language61.5%25.6%
EQ-Bench Creative Writing1258—

Frequently asked questions

Is GPT-4.5 better than Phi-4?

GPT-4.5 is the stronger model overall, scoring 37.2 to 31.2 on the Noometry Index.

Is GPT-4.5 or Phi-4 better for coding?

GPT-4.5 scores higher on coding benchmarks: 42.2 versus 34.4 in the Noometry coding category.

How many benchmarks do GPT-4.5 and Phi-4 share?

29 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Phi-4 has 37.

Related comparisons

Go deeper