Model comparison

Grok 4 vs Step 3.7 Flash

Grok 4 is the stronger model overall, scoring 48.1 to 37.3 on the Noometry Index.

Last verified . 0 shared benchmarks.

Grok 4 xAI

48.1

Rank #56 Confirmed

Step 3.7 Flash StepFun

37.3

Rank #207 Reported

Summary

  • The widest gap is in reasoning, where Grok 4 leads 36.7 to 21.6.
  • Step 3.7 Flash has downloadable open weights; the other is API-only.

Side by side

Grok 4 and Step 3.7 Flash specifications
Grok 4Step 3.7 Flash
ProviderxAIStepFun
Noometry Index48.137.3
Released2025-07-092026-05-29
WeightsProprietaryOpen
Context window—256K
Max output—256K
Input $ / M tokens—$0.18
Output $ / M tokens—$1.11
Results tracked485

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

Category by category

Coding Grok 4 leads

Grok 4: 50.3 (#46), Step 3.7 Flash: 40.0 (#150)

Coding benchmarks
BenchmarkGrok 4Step 3.7 Flash
Aider Polyglot79.6%—
SciCode—40%
WeirdML45.7%—
LMArena Coding1408—
ALE-Bench—694.12

Agentic & Tool Use Not comparable

Grok 4: 32.3 (#68), Step 3.7 Flash: —

Agentic & Tool Use benchmarks
BenchmarkGrok 4Step 3.7 Flash
Terminal-Bench27.2%—
Berkeley Function Calling Leaderboard63%—
GDPval21.1%—
Cybench43%—
DeepResearch Bench47.3%—
BALROG43.6%—
LMArena Search1142—
METR Time Horizons66.6%—

Reasoning Grok 4 leads

Grok 4: 36.7 (#65), Step 3.7 Flash: 21.6 (#219)

Reasoning benchmarks
BenchmarkGrok 4Step 3.7 Flash
ARC-AGI-216%—
SimpleBench60.5%—
Kagi LLM Benchmark73.6%—
NYT Connections (extended)—39.7%
ARC-AGI-166.7%—
CritPt—2.3%
Chess Puzzles28%—
LMArena Hard Prompts1409—
Epoch Capabilities Index146.44—
ForecastBench60.9—

Math Grok 4 leads

Grok 4: 48.4 (#64), Step 3.7 Flash: 42.9 (#82)

Math benchmarks
BenchmarkGrok 4Step 3.7 Flash
MathArena Final-Answer Competitions—68.5%
OTIS Mock AIME 2024-202584%—
Omni-MATH60.3%—
LMArena Math1422—
FrontierMath (Feb 2025 set)19.7%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge Not comparable

Grok 4: 53.8 (#55), Step 3.7 Flash: —

Knowledge benchmarks
BenchmarkGrok 4Step 3.7 Flash
GPQA Diamond87%—
MMLU-Pro85.1%—
Confabulations12.4%—
GPQA (HELM)72.7%—
LMArena Expert1415—

Multimodal Not comparable

Grok 4: 33.7 (#94), Step 3.7 Flash: —

Multimodal benchmarks
BenchmarkGrok 4Step 3.7 Flash
LMArena Vision1210—
GeoBench45%—

Multilingual Not comparable

Grok 4: 51.8 (#103), Step 3.7 Flash: —

Multilingual benchmarks
BenchmarkGrok 4Step 3.7 Flash
LMArena Non-English1403—
LMArena Chinese1427—
LMArena French1418—
LMArena German1429—
LMArena Japanese1394—
LMArena Korean1377—
LMArena Russian1410—
LMArena Spanish1420—

Instruction Following Not comparable

Grok 4: 79.2 (#5), Step 3.7 Flash: —

Instruction Following benchmarks
BenchmarkGrok 4Step 3.7 Flash
IFEval94.9%—
LMArena Instruction Following1387—

Long Context Not comparable

Grok 4: 63.1 (#4), Step 3.7 Flash: —

Long Context benchmarks
BenchmarkGrok 4Step 3.7 Flash
Fiction.LiveBench94.4%—
LMArena Longer Query1409—

Writing & Preference Not comparable

Grok 4: 58.5 (#116), Step 3.7 Flash: —

Writing & Preference benchmarks
BenchmarkGrok 4Step 3.7 Flash
LMArena Text1411—
LMArena Creative Writing1397—
Short-Story Creative Writing76.9%—
WildBench79.7%—
LMArena Multi-Turn1416—

Frequently asked questions

Is Grok 4 better than Step 3.7 Flash?

Grok 4 is the stronger model overall, scoring 48.1 to 37.3 on the Noometry Index.

Is Grok 4 or Step 3.7 Flash better for coding?

Grok 4 scores higher on coding benchmarks: 50.3 versus 40.0 in the Noometry coding category.

How many benchmarks do Grok 4 and Step 3.7 Flash share?

0 benchmarks have published results for both models. Grok 4 has 48 scored results on Noometry and Step 3.7 Flash has 5.

Related comparisons

Go deeper