Noometry

Model comparison

GPT-4.1 mini vs Phi-4

GPT-4.1 mini is the stronger model overall, scoring 33.6 to 31.2 on the Noometry Index. Phi-4 costs 8.0× less per token, which makes it the better buy when GPT-4.1 mini's lead doesn't matter for your workload.

Last verified . 24 shared benchmarks.

GPT-4.1 mini OpenAI

33.6

Rank #240 Confirmed

Phi-4 Microsoft

31.2

Rank #279 Confirmed

Summary

  • They share 24 benchmarks with published results for both. GPT-4.1 mini scores higher in 6 categories and Phi-4 in 3 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in instruction following, where GPT-4.1 mini leads 73.7 to 60.4.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 44.7% for GPT-4.1 mini and 13.8% for Phi-4.
  • Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.40 / $1.60 for GPT-4.1 mini.
  • GPT-4.1 mini accepts more context: 1.05M tokens versus 128K.
  • Phi-4 has downloadable open weights; the other is API-only.

Side by side

GPT-4.1 miniPhi-4
ProviderOpenAIMicrosoft
Noometry Index33.631.2
Released2025-04-142024-12-11
WeightsProprietaryOpen
Context window1.05M128K
Max output33K4K
Input $ / M tokens$0.40$0.07
Output $ / M tokens$1.60$0.14
Results tracked4737

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

Category by category

Coding Phi-4 leads

GPT-4.1 mini: 30.6 (#293), Phi-4: 34.4 (#239)

BenchmarkGPT-4.1 miniPhi-4
BigCodeBench Instruct48.9%45.5%
LMArena Coding13671231
SWE-bench Verified (bash only)23.9%—
Aider Polyglot32.4%—
SciCode40.4%—
WeirdML37.6%—
LiveBench Coding—30.7%
BigCodeBench Complete—55.4%
CadEval16%—

Agentic & Tool Use GPT-4.1 mini leads

GPT-4.1 mini: 33.3 (#55), Phi-4: 22.8 (#128)

BenchmarkGPT-4.1 miniPhi-4
Berkeley Function Calling Leaderboard50.5%28.8%
BALROG—11.6%

Reasoning Phi-4 leads

GPT-4.1 mini: 10.8 (#340), Phi-4: 17.7 (#291)

BenchmarkGPT-4.1 miniPhi-4
Chess Puzzles7%1%
LMArena Hard Prompts13491220
Epoch Capabilities Index135.01130.42
ARC-AGI-20%—
Kagi LLM Benchmark48.6%—
ARC-AGI-13.5%—
CritPt0%—
LiveBench Reasoning—47.8%
Mystery Game Puzzles7%—
DTBench68.8%—
LiveBench Data Analysis—45.2%
LMCA21.1%—
LiveBench—41.6%

Math GPT-4.1 mini leads

GPT-4.1 mini: 24.1 (#270), Phi-4: 20.8 (#285)

BenchmarkGPT-4.1 miniPhi-4
OTIS Mock AIME 2024-202544.7%13.8%
LMArena Math13431246
MATH Level 587.3%64.9%
FrontierMath (Tiers 1-3)6.7%—
Omni-MATH49.1%—
LiveBench Math—42%
FrontierMath (Feb 2025 set)4.5%—

Knowledge GPT-4.1 mini leads

GPT-4.1 mini: 34.7 (#194), Phi-4: 32.6 (#209)

BenchmarkGPT-4.1 miniPhi-4
GPQA Diamond65.8%56.1%
LMArena Expert13381203
SimpleQA Verified12.7%—
MMLU-Pro78.3%—
Confabulations—29.4%
Vectara Hallucination Rate—3.7%
GPQA (HELM)61.4%—
MMLU—84.8%

Multimodal Not comparable

GPT-4.1 mini: 35.8 (#82), Phi-4: —

BenchmarkGPT-4.1 miniPhi-4
LMArena Vision1181—

Multilingual GPT-4.1 mini leads

GPT-4.1 mini: 45.7 (#166), Phi-4: 37.2 (#237)

BenchmarkGPT-4.1 miniPhi-4
LMArena Non-English13181197
LMArena Chinese13291212
LMArena French13581224
LMArena German13511222
LMArena Japanese12901158
LMArena Korean12981151
LMArena Russian13241209
LMArena Spanish13191234

Instruction Following GPT-4.1 mini leads

GPT-4.1 mini: 73.7 (#118), Phi-4: 60.4 (#251)

BenchmarkGPT-4.1 miniPhi-4
LMArena Instruction Following13331201
LiveBench Instruction Following—58.4%
IFEval90.4%—

Long Context Phi-4 leads

GPT-4.1 mini: 31.8 (#275), Phi-4: 36.9 (#226)

BenchmarkGPT-4.1 miniPhi-4
LMArena Longer Query13441217
Fiction.LiveBench44.4%—

Writing & Preference GPT-4.1 mini leads

GPT-4.1 mini: 48.6 (#199), Phi-4: 40.5 (#244)

BenchmarkGPT-4.1 miniPhi-4
LMArena Text13401217
LMArena Creative Writing13001182
LMArena Multi-Turn13541206
Short-Story Creative Writing—62.6%
EQ-Bench Creative Writing1147—
WildBench83.8%—
LiveBench Language—25.6%

Frequently asked questions

Is GPT-4.1 mini better than Phi-4?

GPT-4.1 mini is the stronger model overall, scoring 33.6 to 31.2 on the Noometry Index. Phi-4 costs 8.0× less per token, which makes it the better buy when GPT-4.1 mini's lead doesn't matter for your workload.

Which is cheaper, GPT-4.1 mini or Phi-4?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; GPT-4.1 mini lists at $0.40 and $1.60.

Is GPT-4.1 mini or Phi-4 better for coding?

Phi-4 scores higher on coding benchmarks: 34.4 versus 30.6 in the Noometry coding category.

Which has the bigger context window?

GPT-4.1 mini does, with 1.05M tokens against 128K.

How many benchmarks do GPT-4.1 mini and Phi-4 share?

24 benchmarks have published results for both models. GPT-4.1 mini has 47 scored results on Noometry and Phi-4 has 37.

Related comparisons

Go deeper