Noometry

Model comparison

Granite 4.2 8B vs Phi-4

Granite 4.2 8B is the stronger model overall, scoring 40.0 to 30.9 on the Noometry Index.

Last verified . 11 shared benchmarks.

Granite 4.2 8B IBM

40.0

Rank #143 Confirmed

Phi-4 Microsoft

30.9

Rank #269 Confirmed

Summary

  • They share 11 benchmarks with published results for both. Granite 4.2 8B scores higher in 7 categories and Phi-4 in 0 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where Granite 4.2 8B leads 49.8 to 40.7.
  • Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.06 / $0.25 for Granite 4.2 8B.
  • Granite 4.2 8B accepts more context: 131K tokens versus 128K.

Side by side

Granite 4.2 8BPhi-4
ProviderIBMMicrosoft
Noometry Index40.030.9
Released—2024-12-11
WeightsOpenOpen
Context window131K128K
Max output118K4K
Input $ / M tokens$0.06$0.07
Output $ / M tokens$0.25$0.14
Results tracked1137

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

Category by category

Coding Granite 4.2 8B leads

Granite 4.2 8B: 39.9 (#138), Phi-4: 33.8 (#237)

BenchmarkGranite 4.2 8BPhi-4
LMArena Coding13791231
BigCodeBench Instruct—45.5%
LiveBench Coding—30.7%
BigCodeBench Complete—55.4%

Agentic & Tool Use Not comparable

Granite 4.2 8B: —, Phi-4: 22.6 (#120)

BenchmarkGranite 4.2 8BPhi-4
Berkeley Function Calling Leaderboard—28.8%
BALROG—11.6%

Reasoning Granite 4.2 8B leads

Granite 4.2 8B: 25.5 (#130), Phi-4: 17.0 (#287)

BenchmarkGranite 4.2 8BPhi-4
LMArena Hard Prompts13271220
Chess Puzzles—1%
LiveBench Reasoning—47.8%
LiveBench Data Analysis—45.2%
Epoch Capabilities Index—130.43
LiveBench—41.6%

Math Not comparable

Granite 4.2 8B: —, Phi-4: 19.6 (#279)

BenchmarkGranite 4.2 8BPhi-4
OTIS Mock AIME 2024-2025—13.8%
LiveBench Math—42%
LMArena Math—1246
MATH Level 5—64.9%

Knowledge Granite 4.2 8B leads

Granite 4.2 8B: 37.7 (#144), Phi-4: 32.5 (#204)

BenchmarkGranite 4.2 8BPhi-4
LMArena Expert13801203
GPQA Diamond—56.1%
Confabulations—29.4%
Vectara Hallucination Rate—3.7%
MMLU—84.8%

Multilingual Granite 4.2 8B leads

Granite 4.2 8B: 44.5 (#173), Phi-4: 37.4 (#232)

BenchmarkGranite 4.2 8BPhi-4
LMArena Non-English12991197
LMArena Chinese13611211
LMArena Russian12831209
LMArena French—1223
LMArena German—1222
LMArena Japanese—1158
LMArena Korean—1151
LMArena Spanish—1234

Instruction Following Granite 4.2 8B leads

Granite 4.2 8B: 68.7 (#183), Phi-4: 60.6 (#246)

BenchmarkGranite 4.2 8BPhi-4
LMArena Instruction Following12981201
LiveBench Instruction Following—58.4%

Long Context Granite 4.2 8B leads

Granite 4.2 8B: 40.2 (#154), Phi-4: 36.9 (#221)

BenchmarkGranite 4.2 8BPhi-4
LMArena Longer Query13231217

Writing & Preference Granite 4.2 8B leads

Granite 4.2 8B: 49.8 (#183), Phi-4: 40.7 (#239)

BenchmarkGranite 4.2 8BPhi-4
LMArena Text13191217
LMArena Creative Writing12381182
LMArena Multi-Turn12991206
Short-Story Creative Writing—62.6%
LiveBench Language—25.6%

Frequently asked questions

Is Granite 4.2 8B better than Phi-4?

Granite 4.2 8B is the stronger model overall, scoring 40.0 to 30.9 on the Noometry Index.

Which is cheaper, Granite 4.2 8B or Phi-4?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Granite 4.2 8B lists at $0.06 and $0.25.

Is Granite 4.2 8B or Phi-4 better for coding?

Granite 4.2 8B scores higher on coding benchmarks: 39.9 versus 33.8 in the Noometry coding category.

Which has the bigger context window?

Granite 4.2 8B does, with 131K tokens against 128K.

How many benchmarks do Granite 4.2 8B and Phi-4 share?

11 benchmarks have published results for both models. Granite 4.2 8B has 11 scored results on Noometry and Phi-4 has 37.

Related comparisons

Go deeper