Noometry

Model comparison

Gemini 2.5 Pro vs Phi-4

Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 31.2 on the Noometry Index. Phi-4 costs 39× less per token, which makes it the better buy when Gemini 2.5 Pro's lead doesn't matter for your workload.

Last verified . 33 shared benchmarks.

Gemini 2.5 Pro Google

45.0

Rank #75 Confirmed

Phi-4 Microsoft

31.2

Rank #279 Confirmed

Summary

  • They share 33 benchmarks with published results for both. Gemini 2.5 Pro scores higher in 9 categories and Phi-4 in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in knowledge, where Gemini 2.5 Pro leads 56.0 to 32.6.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 84.7% for Gemini 2.5 Pro and 13.8% for Phi-4.
  • Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $1.25 / $10 for Gemini 2.5 Pro.
  • Gemini 2.5 Pro accepts more context: 1.05M tokens versus 128K.
  • Phi-4 has downloadable open weights; the other is API-only.

Side by side

Gemini 2.5 ProPhi-4
ProviderGoogleMicrosoft
Noometry Index45.031.2
Released2025-03-252024-12-11
WeightsProprietaryOpen
Context window1.05M128K
Max output66K4K
Input $ / M tokens$1.25$0.07
Output $ / M tokens$10$0.14
Results tracked7837

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

Category by category

Coding Gemini 2.5 Pro leads

Gemini 2.5 Pro: 42.4 (#101), Phi-4: 34.4 (#239)

BenchmarkGemini 2.5 ProPhi-4
LiveBench Coding85.9%30.7%
LMArena Coding14521231
SWE-bench Verified57.6%—
SWE-bench Verified (bash only)53.6%—
Aider Polyglot83.1%—
LMArena WebDev1227—
SciCode42.8%—
GSO3.9%—
WeirdML54%—
BigCodeBench Instruct—45.5%
BigCodeBench Complete—55.4%
CadEval64%—
ALE-Bench785.52—
AlgoTune1.51—

Agentic & Tool Use Gemini 2.5 Pro leads

Gemini 2.5 Pro: 29.2 (#88), Phi-4: 22.8 (#128)

BenchmarkGemini 2.5 ProPhi-4
BALROG43.3%11.6%
Terminal-Bench32.6%—
Berkeley Function Calling Leaderboard—28.8%
GDPval23.3%—
Remote Labor Index0.8%—
TheAgentCompany30.3%—
τ²-bench Banking13.7%—
DeepResearch Bench42.8%—
LMArena Search1142—
METR Time Horizons55.4%—
Vending-Bench 2573.64—

Reasoning Gemini 2.5 Pro leads

Gemini 2.5 Pro: 28.8 (#99), Phi-4: 17.7 (#291)

BenchmarkGemini 2.5 ProPhi-4
Chess Puzzles20%1%
LiveBench Reasoning89.8%47.8%
LMArena Hard Prompts14551220
LiveBench Data Analysis79.9%45.2%
Epoch Capabilities Index145.32130.42
LiveBench82.3%41.6%
ARC-AGI-24.9%—
SimpleBench62.4%—
Kagi LLM Benchmark70.3%—
ARC-AGI-141%—
CritPt2%—
EnigmaEval5.6%—
DTBench82.4%—
LMCA34.8%—
ForecastBench61.3—

Math Gemini 2.5 Pro leads

Gemini 2.5 Pro: 32.5 (#213), Phi-4: 20.8 (#285)

BenchmarkGemini 2.5 ProPhi-4
OTIS Mock AIME 2024-202584.7%13.8%
LiveBench Math90.2%42%
LMArena Math14501246
MATH Level 595.9%64.9%
FrontierMath (Tiers 1-3)24.6%—
FrontierMath Tier 40%—
Omni-MATH41.6%—
FrontierMath (Feb 2025 set)14.1%—
FrontierMath Tier 4 (v1)4.2%—

Knowledge Gemini 2.5 Pro leads

Gemini 2.5 Pro: 56.0 (#46), Phi-4: 32.6 (#209)

BenchmarkGemini 2.5 ProPhi-4
GPQA Diamond85.3%56.1%
Confabulations10.6%29.4%
Vectara Hallucination Rate7%3.7%
LMArena Expert14521203
Humanity's Last Exam21.6%—
MMLU-Pro86.3%—
GPQA (HELM)74.9%—
MMLU—84.8%

Multimodal Not comparable

Gemini 2.5 Pro: 45.2 (#18), Phi-4: —

BenchmarkGemini 2.5 ProPhi-4
LMArena Vision1263—
GeoBench86%—
VPCT48%—
LMArena Document1421—
SpatialViz-Bench44.7%—

Multilingual Gemini 2.5 Pro leads

Gemini 2.5 Pro: 55.3 (#31), Phi-4: 37.2 (#237)

BenchmarkGemini 2.5 ProPhi-4
LMArena Non-English14511197
LMArena Chinese15071212
LMArena French14721224
LMArena German14871222
LMArena Japanese14611158
LMArena Korean14341151
LMArena Russian14611209
LMArena Spanish14731234

Instruction Following Gemini 2.5 Pro leads

Gemini 2.5 Pro: 75.0 (#75), Phi-4: 60.4 (#251)

BenchmarkGemini 2.5 ProPhi-4
LiveBench Instruction Following80.6%58.4%
LMArena Instruction Following14371201
IFEval84%—

Long Context Gemini 2.5 Pro leads

Gemini 2.5 Pro: 59.8 (#5), Phi-4: 36.9 (#226)

BenchmarkGemini 2.5 ProPhi-4
LMArena Longer Query14491217
Fiction.LiveBench91.7%—

Writing & Preference Gemini 2.5 Pro leads

Gemini 2.5 Pro: 63.7 (#62), Phi-4: 40.5 (#244)

BenchmarkGemini 2.5 ProPhi-4
LMArena Text14581217
LMArena Creative Writing14541182
Short-Story Creative Writing83.8%62.6%
LMArena Multi-Turn14531206
LiveBench Language67.8%25.6%
EQ-Bench Creative Writing1421—
WildBench85.7%—

Frequently asked questions

Is Gemini 2.5 Pro better than Phi-4?

Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 31.2 on the Noometry Index. Phi-4 costs 39× less per token, which makes it the better buy when Gemini 2.5 Pro's lead doesn't matter for your workload.

Which is cheaper, Gemini 2.5 Pro or Phi-4?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Gemini 2.5 Pro lists at $1.25 and $10.

Is Gemini 2.5 Pro or Phi-4 better for coding?

Gemini 2.5 Pro scores higher on coding benchmarks: 42.4 versus 34.4 in the Noometry coding category.

Which has the bigger context window?

Gemini 2.5 Pro does, with 1.05M tokens against 128K.

How many benchmarks do Gemini 2.5 Pro and Phi-4 share?

33 benchmarks have published results for both models. Gemini 2.5 Pro has 78 scored results on Noometry and Phi-4 has 37.

Related comparisons

Go deeper