Noometry

Model comparison

Phi-4 vs Qwen3 8B

Qwen3 8B is the stronger model overall, scoring 33.7 to 31.2 on the Noometry Index. Phi-4 costs 3.5× less per token, which makes it the better buy when Qwen3 8B's lead doesn't matter for your workload.

Last verified . 6 shared benchmarks.

Phi-4 Microsoft

31.2

Rank #279 Confirmed

Qwen3 8B Alibaba (Qwen)

33.7

Rank #238 Confirmed

Summary

  • They share 6 benchmarks with published results for both. Phi-4 scores higher in 2 categories and Qwen3 8B in 4 categories; 4 gaps are clear of the uncertainty.
  • The widest gap is in math, where Qwen3 8B leads 34.9 to 20.8.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 13.8% for Phi-4 and 56.1% for Qwen3 8B.
  • Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.18 / $0.70 for Qwen3 8B.
  • Qwen3 8B accepts more context: 131K tokens versus 128K.

Side by side

Phi-4Qwen3 8B
ProviderMicrosoftAlibaba (Qwen)
Noometry Index31.233.7
Released2024-12-112025-04
WeightsOpenOpen
Context window128K131K
Max output4K8K
Input $ / M tokens$0.07$0.18
Output $ / M tokens$0.14$0.70
Results tracked3711

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

Category by category

Coding Too close to call

Phi-4: 34.4 (#239), Qwen3 8B: 34.0 (#248)

BenchmarkPhi-4Qwen3 8B
SciCode—22.6%
BigCodeBench Instruct45.5%—
LiveBench Coding30.7%—
LMArena Coding1231—
BigCodeBench Complete55.4%—

Agentic & Tool Use Qwen3 8B leads

Phi-4: 22.8 (#128), Qwen3 8B: 30.2 (#78)

BenchmarkPhi-4Qwen3 8B
Berkeley Function Calling Leaderboard28.8%42.6%
BALROG11.6%—

Reasoning Phi-4 leads

Phi-4: 17.7 (#291), Qwen3 8B: 16.6 (#303)

BenchmarkPhi-4Qwen3 8B
Chess Puzzles1%5%
Epoch Capabilities Index130.42136.17
CritPt—0%
LiveBench Reasoning47.8%—
LMArena Hard Prompts1220—
DTBench—59.7%
LiveBench Data Analysis45.2%—
LMCA—8.8%
LiveBench41.6%—

Math Qwen3 8B leads

Phi-4: 20.8 (#285), Qwen3 8B: 34.9 (#191)

BenchmarkPhi-4Qwen3 8B
OTIS Mock AIME 2024-202513.8%56.1%
LiveBench Math42%—
LMArena Math1246—
MATH Level 564.9%—

Knowledge Qwen3 8B leads

Phi-4: 32.6 (#209), Qwen3 8B: 36.1 (#173)

BenchmarkPhi-4Qwen3 8B
GPQA Diamond56.1%56.8%
Vectara Hallucination Rate3.7%4.8%
Confabulations29.4%—
LMArena Expert1203—
MMLU84.8%—

Multilingual Not comparable

Phi-4: 37.2 (#237), Qwen3 8B: —

BenchmarkPhi-4Qwen3 8B
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), Qwen3 8B: —

BenchmarkPhi-4Qwen3 8B
LiveBench Instruction Following58.4%—
LMArena Instruction Following1201—

Long Context Too close to call

Phi-4: 36.9 (#226), Qwen3 8B: 37.9 (#210)

BenchmarkPhi-4Qwen3 8B
Fiction.LiveBench—62.1%
LMArena Longer Query1217—

Writing & Preference Not comparable

Phi-4: 40.5 (#244), Qwen3 8B: —

BenchmarkPhi-4Qwen3 8B
LMArena Text1217—
LMArena Creative Writing1182—
Short-Story Creative Writing62.6%—
LMArena Multi-Turn1206—
LiveBench Language25.6%—

Frequently asked questions

Is Phi-4 better than Qwen3 8B?

Qwen3 8B is the stronger model overall, scoring 33.7 to 31.2 on the Noometry Index. Phi-4 costs 3.5× less per token, which makes it the better buy when Qwen3 8B's lead doesn't matter for your workload.

Which is cheaper, Phi-4 or Qwen3 8B?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Qwen3 8B lists at $0.18 and $0.70.

Is Phi-4 or Qwen3 8B better for coding?

They score almost the same on coding (34.4 vs 34.0); test both on your own repository before choosing.

Which has the bigger context window?

Qwen3 8B does, with 131K tokens against 128K.

How many benchmarks do Phi-4 and Qwen3 8B share?

6 benchmarks have published results for both models. Phi-4 has 37 scored results on Noometry and Qwen3 8B has 11.

Related comparisons

Go deeper