Noometry

Model comparison

GPT-4 Turbo vs Step 5 Preview

Step 5 Preview is the stronger model overall, scoring 47.9 to 30.5 on the Noometry Index.

Last verified . 14 shared benchmarks.

GPT-4 Turbo OpenAI

30.5

Rank #292 Confirmed

Step 5 Preview StepFun

47.9

Rank #58 Confirmed

Summary

  • They share 14 benchmarks with published results for both. GPT-4 Turbo scores higher in 0 categories and Step 5 Preview in 9 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in math, where Step 5 Preview leads 46.1 to 9.0.
  • Step 5 Preview is cheaper at $1 / $2.70 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
  • Step 5 Preview accepts more context: 1.02M tokens versus 128K.

Side by side

GPT-4 TurboStep 5 Preview
ProviderOpenAIStepFun
Noometry Index30.547.9
Released2023-11-062026-09-16
WeightsProprietaryProprietary
Context window128K1.02M
Max output4K66K
Input $ / M tokens$10$1
Output $ / M tokens$30$2.70
Results tracked3618

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

Category by category

Coding Step 5 Preview leads

GPT-4 Turbo: 33.8 (#249), Step 5 Preview: 51.8 (#36)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Coding12681480
LMArena WebDev—1564
SciCode—58.9%
WeirdML18%—
BigCodeBench Instruct48.2%—
BigCodeBench Complete58.2%—
HumanEval+86.6%—
MBPP+73.3%—

Agentic & Tool Use Not comparable

GPT-4 Turbo: —, Step 5 Preview: —

BenchmarkGPT-4 TurboStep 5 Preview
METR Time Horizons36.7%—

Reasoning Step 5 Preview leads

GPT-4 Turbo: 15.3 (#317), Step 5 Preview: 40.0 (#57)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Hard Prompts12511465
SimpleBench25.1%—
CritPt—20.9%
Chess Puzzles6%—
DTBench61.6%—
LMCA9.8%—
Epoch Capabilities Index127.25—
ForecastBench59.4—

Math Step 5 Preview leads

GPT-4 Turbo: 9.0 (#322), Step 5 Preview: 46.1 (#72)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Math12721470
FrontierMath (Tiers 1-3)0.7%—
OTIS Mock AIME 2024-20256.7%—
ProofBench—42%
MATH Level 546.7%—

Knowledge Step 5 Preview leads

GPT-4 Turbo: 24.3 (#268), Step 5 Preview: 41.2 (#112)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Expert12231470
GPQA Diamond46.6%—
Confabulations28.4%—
MMLU81.3%—

Multimodal Step 5 Preview leads

GPT-4 Turbo: 30.6 (#110), Step 5 Preview: 41.0 (#41)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Vision10901267

Multilingual Step 5 Preview leads

GPT-4 Turbo: 40.5 (#216), Step 5 Preview: 53.8 (#54)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Non-English12451432
LMArena Chinese12421519
LMArena Russian12591436
LMArena Spanish12601439
LMArena French1276—
LMArena German1259—
LMArena Japanese1194—
LMArena Korean1187—

Instruction Following Step 5 Preview leads

GPT-4 Turbo: 65.8 (#216), Step 5 Preview: 76.0 (#50)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Instruction Following12491444

Long Context Step 5 Preview leads

GPT-4 Turbo: 38.0 (#206), Step 5 Preview: 44.5 (#64)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Longer Query12541455

Writing & Preference Step 5 Preview leads

GPT-4 Turbo: 47.7 (#206), Step 5 Preview: 62.8 (#71)

BenchmarkGPT-4 TurboStep 5 Preview
LMArena Text12721442
LMArena Creative Writing12691410
LMArena Multi-Turn12671447

Frequently asked questions

Is GPT-4 Turbo better than Step 5 Preview?

Step 5 Preview is the stronger model overall, scoring 47.9 to 30.5 on the Noometry Index.

Which is cheaper, GPT-4 Turbo or Step 5 Preview?

Step 5 Preview is cheaper. It lists at $1 per million input tokens and $2.70 per million output tokens; GPT-4 Turbo lists at $10 and $30.

Is GPT-4 Turbo or Step 5 Preview better for coding?

Step 5 Preview scores higher on coding benchmarks: 51.8 versus 33.8 in the Noometry coding category.

Which has the bigger context window?

Step 5 Preview does, with 1.02M tokens against 128K.

How many benchmarks do GPT-4 Turbo and Step 5 Preview share?

14 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and Step 5 Preview has 18.

Related comparisons

Go deeper