Noometry

Model comparison

GPT-4.1 vs Step 3

Step 3 is the stronger model overall, scoring 40.5 to 35.9 on the Noometry Index.

Last verified . 17 shared benchmarks.

GPT-4.1 OpenAI

35.9

Rank #219 Confirmed

Step 3 StepFun

40.5

Rank #149 Confirmed

Summary

  • They share 17 benchmarks with published results for both. GPT-4.1 scores higher in 5 categories and Step 3 in 4 categories; 6 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Step 3 leads 28.4 to 11.7.
  • The biggest single-benchmark swing is Kagi LLM Benchmark: 52.3% for GPT-4.1 and 62.3% for Step 3.
  • Step 3 has downloadable open weights; the other is API-only.

Side by side

GPT-4.1Step 3
ProviderOpenAIStepFun
Noometry Index35.940.5
Released2025-04-14—
WeightsProprietaryOpen
Context window1.05M—
Max output33K—
Input $ / M tokens$2—
Output $ / M tokens$8—
Results tracked5217

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

Category by category

Coding Step 3 leads

GPT-4.1: 34.4 (#238), Step 3: 40.1 (#147)

BenchmarkGPT-4.1Step 3
LMArena Coding13911367
SWE-bench Verified48.5%—
SWE-bench Verified (bash only)39.6%—
Aider Polyglot52.4%—
WeirdML39%—
CadEval42%—
ALE-Bench558.1—

Agentic & Tool Use Not comparable

GPT-4.1: 34.7 (#43), Step 3: —

BenchmarkGPT-4.1Step 3
Berkeley Function Calling Leaderboard54%—

Reasoning Step 3 leads

GPT-4.1: 11.7 (#339), Step 3: 28.4 (#105)

BenchmarkGPT-4.1Step 3
Kagi LLM Benchmark52.3%62.3%
LMArena Hard Prompts13841355
ARC-AGI-20.4%—
SimpleBench27%—
ARC-AGI-15.5%—
Chess Puzzles6%—
EnigmaEval2.2%—
DTBench68.3%—
LMCA25.6%—
Epoch Capabilities Index136.78—
ForecastBench61.5—

Math Step 3 leads

GPT-4.1: 22.3 (#280), Step 3: 37.6 (#148)

BenchmarkGPT-4.1Step 3
LMArena Math13701366
FrontierMath (Tiers 1-3)6%—
OTIS Mock AIME 2024-202538.3%—
Omni-MATH47.1%—
MATH Level 583%—
FrontierMath (Feb 2025 set)5.5%—
FrontierMath Tier 4 (v1)0%—

Knowledge Too close to call

GPT-4.1: 37.1 (#160), Step 3: 36.8 (#164)

BenchmarkGPT-4.1Step 3
LMArena Expert13641333
GPQA Diamond66.9%—
Humanity's Last Exam5.4%—
SimpleQA Verified31.1%—
MMLU-Pro81.1%—
Vectara Hallucination Rate5.6%—
GPQA (HELM)65.9%—

Multimodal GPT-4.1 leads

GPT-4.1: 38.2 (#67), Step 3: 35.5 (#86)

BenchmarkGPT-4.1Step 3
LMArena Vision12111177
GeoBench72%—

Multilingual GPT-4.1 leads

GPT-4.1: 49.4 (#133), Step 3: 46.3 (#159)

BenchmarkGPT-4.1Step 3
LMArena Non-English13701327
LMArena Chinese13821397
LMArena German13811371
LMArena Korean13391269
LMArena Russian13771331
LMArena Spanish13761371
LMArena French1382—
LMArena Japanese1319—

Instruction Following Too close to call

GPT-4.1: 71.3 (#153), Step 3: 70.4 (#164)

BenchmarkGPT-4.1Step 3
LMArena Instruction Following13671332
IFEval83.8%—

Long Context Too close to call

GPT-4.1: 40.0 (#163), Step 3: 40.3 (#157)

BenchmarkGPT-4.1Step 3
LMArena Longer Query13851326
Fiction.LiveBench63.9%—

Writing & Preference GPT-4.1 leads

GPT-4.1: 57.6 (#125), Step 3: 54.3 (#151)

BenchmarkGPT-4.1Step 3
LMArena Text13831350
LMArena Creative Writing13631321
LMArena Multi-Turn13981341
EQ-Bench Creative Writing1420—
WildBench85.4%—

Frequently asked questions

Is GPT-4.1 better than Step 3?

Step 3 is the stronger model overall, scoring 40.5 to 35.9 on the Noometry Index.

Is GPT-4.1 or Step 3 better for coding?

Step 3 scores higher on coding benchmarks: 40.1 versus 34.4 in the Noometry coding category.

How many benchmarks do GPT-4.1 and Step 3 share?

17 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Step 3 has 17.

Related comparisons

Go deeper