Noometry

Model comparison

Phi-4 vs Qwen2.5 32B Instruct

Phi-4 is the stronger model overall, scoring 31.2 to 30.1 on the Noometry Index.

Last verified . 7 shared benchmarks.

Phi-4 Microsoft

31.2

Rank #279 Confirmed

Qwen2.5 32B Instruct Alibaba (Qwen)

30.1

Rank #297 Confirmed

Summary

  • They share 7 benchmarks with published results for both. Phi-4 scores higher in 2 categories and Qwen2.5 32B Instruct in 2 categories; 4 gaps are clear of the uncertainty.
  • The widest gap is in knowledge, where Phi-4 leads 32.6 to 24.9.
  • The biggest single-benchmark swing is GPQA Diamond: 56.1% for Phi-4 and 46.1% for Qwen2.5 32B Instruct.
  • Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.70 / $2.80 for Qwen2.5 32B Instruct.
  • Qwen2.5 32B Instruct accepts more context: 131K tokens versus 128K.

Side by side

Phi-4Qwen2.5 32B Instruct
ProviderMicrosoftAlibaba (Qwen)
Noometry Index31.230.1
Released2024-12-112024-09
WeightsOpenOpen
Context window128K131K
Max output4K8K
Input $ / M tokens$0.07$0.70
Output $ / M tokens$0.14$2.80
Results tracked377

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

Category by category

Coding Qwen2.5 32B Instruct leads

Phi-4: 34.4 (#239), Qwen2.5 32B Instruct: 38.7 (#169)

BenchmarkPhi-4Qwen2.5 32B Instruct
BigCodeBench Instruct45.5%45%
BigCodeBench Complete55.4%52.3%
LiveBench Coding30.7%—
LMArena Coding1231—

Agentic & Tool Use Not comparable

Phi-4: 22.8 (#128), Qwen2.5 32B Instruct: —

BenchmarkPhi-4Qwen2.5 32B Instruct
Berkeley Function Calling Leaderboard28.8%—
BALROG11.6%—

Reasoning Qwen2.5 32B Instruct leads

Phi-4: 17.7 (#291), Qwen2.5 32B Instruct: 19.2 (#266)

BenchmarkPhi-4Qwen2.5 32B Instruct
Chess Puzzles1%0%
Epoch Capabilities Index130.42128.52
LiveBench Reasoning47.8%—
LMArena Hard Prompts1220—
LiveBench Data Analysis45.2%—
LiveBench41.6%—

Math Phi-4 leads

Phi-4: 20.8 (#285), Qwen2.5 32B Instruct: 16.2 (#296)

BenchmarkPhi-4Qwen2.5 32B Instruct
OTIS Mock AIME 2024-202513.8%7.4%
MATH Level 564.9%56.1%
LiveBench Math42%—
LMArena Math1246—

Knowledge Phi-4 leads

Phi-4: 32.6 (#209), Qwen2.5 32B Instruct: 24.9 (#266)

BenchmarkPhi-4Qwen2.5 32B Instruct
GPQA Diamond56.1%46.1%
Confabulations29.4%—
Vectara Hallucination Rate3.7%—
LMArena Expert1203—
MMLU84.8%—

Multilingual Not comparable

Phi-4: 37.2 (#237), Qwen2.5 32B Instruct: —

BenchmarkPhi-4Qwen2.5 32B Instruct
LMArena Non-English1197—
LMArena Chinese1212—
LMArena French1224—
LMArena German1222—
LMArena Japanese1158—
LMArena Korean1151—
LMArena Russian1209—
LMArena Spanish1234—

Instruction Following Not comparable

Phi-4: 60.4 (#251), Qwen2.5 32B Instruct: —

BenchmarkPhi-4Qwen2.5 32B Instruct
LiveBench Instruction Following58.4%—
LMArena Instruction Following1201—

Long Context Not comparable

Phi-4: 36.9 (#226), Qwen2.5 32B Instruct: —

BenchmarkPhi-4Qwen2.5 32B Instruct
LMArena Longer Query1217—

Writing & Preference Not comparable

Phi-4: 40.5 (#244), Qwen2.5 32B Instruct: —

BenchmarkPhi-4Qwen2.5 32B Instruct
LMArena Text1217—
LMArena Creative Writing1182—
Short-Story Creative Writing62.6%—
LMArena Multi-Turn1206—
LiveBench Language25.6%—

Frequently asked questions

Is Phi-4 better than Qwen2.5 32B Instruct?

Phi-4 is the stronger model overall, scoring 31.2 to 30.1 on the Noometry Index.

Which is cheaper, Phi-4 or Qwen2.5 32B Instruct?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Qwen2.5 32B Instruct lists at $0.70 and $2.80.

Is Phi-4 or Qwen2.5 32B Instruct better for coding?

Qwen2.5 32B Instruct scores higher on coding benchmarks: 38.7 versus 34.4 in the Noometry coding category.

Which has the bigger context window?

Qwen2.5 32B Instruct does, with 131K tokens against 128K.

How many benchmarks do Phi-4 and Qwen2.5 32B Instruct share?

7 benchmarks have published results for both models. Phi-4 has 37 scored results on Noometry and Qwen2.5 32B Instruct has 7.

Related comparisons

Go deeper