Model comparison

Phi 3 Mini 4k Instruct vs Step 2 16k Exp 202412

Step 2 16k Exp 202412 is the stronger model overall, scoring 38.6 to 27.4 on the Noometry Index.

Last verified . 12 shared benchmarks.

Phi 3 Mini 4k Instruct Microsoft

27.4

Rank #324 Confirmed

Step 2 16k Exp 202412 StepFun

38.6

Rank #173 Confirmed

Summary

  • They share 12 benchmarks with published results for both. Phi 3 Mini 4k Instruct scores higher in 0 categories and Step 2 16k Exp 202412 in 8 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where Step 2 16k Exp 202412 leads 52.5 to 27.7.
  • Phi 3 Mini 4k Instruct has downloadable open weights; the other is API-only.

Side by side

Phi 3 Mini 4k Instruct and Step 2 16k Exp 202412 specifications
Phi 3 Mini 4k InstructStep 2 16k Exp 202412
ProviderMicrosoftStepFun
Noometry Index27.438.6
Released2024-04-23—
WeightsOpenProprietary
Context window——
Max output——
Input $ / M tokens——
Output $ / M tokens——
Results tracked3512

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

Category by category

Coding Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 26.1 (#319), Step 2 16k Exp 202412: 37.9 (#170)

Coding benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Coding10931316
LiveBench Coding15.5%—
HumanEval+59.1%—
MBPP+54.2%—

Reasoning Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 13.6 (#326), Step 2 16k Exp 202412: 24.8 (#140)

Reasoning benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Hard Prompts10721299
Chess Puzzles0%—
LiveBench Reasoning26.8%—
LiveBench Data Analysis34.7%—
Adversarial NLI52.8%—
BIG-Bench Hard71.7%—
HellaSwag76.7%—
LiveBench22.4%—
WinoGrande70.8%—

Math Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 24.5 (#258), Step 2 16k Exp 202412: 33.6 (#171)

Math benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Math11111305
LiveBench Math15.7%—

Knowledge Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 27.9 (#241), Step 2 16k Exp 202412: 34.6 (#187)

Knowledge benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Expert10451279
ARC (AI2) Challenge84.9%—
MMLU68.8%—
OpenBookQA88%—
TriviaQA64%—

Multilingual Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 26.5 (#275), Step 2 16k Exp 202412: 43.9 (#176)

Multilingual benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Non-English10211290
LMArena Chinese10211331
LMArena Russian10221324
LMArena French1075—
LMArena German1044—
LMArena Japanese935—
LMArena Korean905—
LMArena Spanish1084—

Instruction Following Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 48.0 (#298), Step 2 16k Exp 202412: 68.1 (#187)

Instruction Following benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Instruction Following10531287
LiveBench Instruction Following39.1%—

Long Context Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 31.7 (#271), Step 2 16k Exp 202412: 39.7 (#165)

Long Context benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Longer Query10441307

Writing & Preference Step 2 16k Exp 202412 leads

Phi 3 Mini 4k Instruct: 27.7 (#295), Step 2 16k Exp 202412: 52.5 (#168)

Writing & Preference benchmarks
BenchmarkPhi 3 Mini 4k InstructStep 2 16k Exp 202412
LMArena Text10731321
LMArena Creative Writing10371328
LMArena Multi-Turn10181292
LiveBench Language9.2%—

Frequently asked questions

Is Phi 3 Mini 4k Instruct better than Step 2 16k Exp 202412?

Step 2 16k Exp 202412 is the stronger model overall, scoring 38.6 to 27.4 on the Noometry Index.

Is Phi 3 Mini 4k Instruct or Step 2 16k Exp 202412 better for coding?

Step 2 16k Exp 202412 scores higher on coding benchmarks: 37.9 versus 26.1 in the Noometry coding category.

How many benchmarks do Phi 3 Mini 4k Instruct and Step 2 16k Exp 202412 share?

12 benchmarks have published results for both models. Phi 3 Mini 4k Instruct has 35 scored results on Noometry and Step 2 16k Exp 202412 has 12.

Related comparisons

Go deeper