Noometry

Model comparison

GPT-4 vs Step 3.7 Flash

Step 3.7 Flash is the stronger model overall, scoring 37.3 to 29.1 on the Noometry Index.

Last verified . 0 shared benchmarks.

GPT-4 OpenAI

29.1

Rank #316 Confirmed

Step 3.7 Flash StepFun

37.3

Rank #207 Reported

Summary

  • The widest gap is in math, where Step 3.7 Flash leads 42.9 to 10.8.
  • Step 3.7 Flash is cheaper at $0.18 / $1.11 per million input/output tokens, against $30 / $60 for GPT-4.
  • Step 3.7 Flash accepts more context: 256K tokens versus 8K.
  • Step 3.7 Flash has downloadable open weights; the other is API-only.

Side by side

GPT-4Step 3.7 Flash
ProviderOpenAIStepFun
Noometry Index29.137.3
Released2023-03-142026-05-29
WeightsProprietaryOpen
Context window8K256K
Max output8K256K
Input $ / M tokens$30$0.18
Output $ / M tokens$60$1.11
Results tracked385

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

Category by category

Coding Step 3.7 Flash leads

GPT-4: 31.6 (#283), Step 3.7 Flash: 40.0 (#150)

BenchmarkGPT-4Step 3.7 Flash
SciCode—40%
WeirdML12.4%—
BigCodeBench Instruct46%—
LMArena Coding1254—
BigCodeBench Complete57.2%—
ALE-Bench—694.12
HumanEval+79.3%—

Agentic & Tool Use Not comparable

GPT-4: —, Step 3.7 Flash: —

BenchmarkGPT-4Step 3.7 Flash
METR Time Horizons36.1%—

Reasoning Step 3.7 Flash leads

GPT-4: 17.8 (#289), Step 3.7 Flash: 21.6 (#219)

BenchmarkGPT-4Step 3.7 Flash
NYT Connections (extended)—39.7%
CritPt—2.3%
Chess Puzzles4%—
LMArena Hard Prompts1241—
Mystery Game Puzzles12%—
DTBench62.7%—
LMCA17.1%—
BIG-Bench Hard75.1%—
Epoch Capabilities Index125.89—
ForecastBench57.8—
HellaSwag95.3%—
WinoGrande87.5%—

Math Step 3.7 Flash leads

GPT-4: 10.8 (#309), Step 3.7 Flash: 42.9 (#82)

BenchmarkGPT-4Step 3.7 Flash
MathArena Final-Answer Competitions—68.5%
OTIS Mock AIME 2024-20251.1%—
LMArena Math1269—
MATH Level 523%—
GSM8K92%—

Knowledge Not comparable

GPT-4: 18.4 (#282), Step 3.7 Flash: —

BenchmarkGPT-4Step 3.7 Flash
GPQA Diamond35.7%—
LMArena Expert1211—
MMLU86.4%—
TriviaQA84.8%—

Multilingual Not comparable

GPT-4: 40.6 (#215), Step 3.7 Flash: —

BenchmarkGPT-4Step 3.7 Flash
LMArena Non-English1246—
LMArena Chinese1242—
LMArena French1283—
LMArena German1251—
LMArena Japanese1209—
LMArena Korean1184—
LMArena Russian1251—
LMArena Spanish1261—

Instruction Following Not comparable

GPT-4: 65.3 (#222), Step 3.7 Flash: —

BenchmarkGPT-4Step 3.7 Flash
LMArena Instruction Following1241—

Long Context Not comparable

GPT-4: 37.7 (#212), Step 3.7 Flash: —

BenchmarkGPT-4Step 3.7 Flash
LMArena Longer Query1244—

Writing & Preference Not comparable

GPT-4: 34.9 (#268), Step 3.7 Flash: —

BenchmarkGPT-4Step 3.7 Flash
LMArena Text1263—
LMArena Creative Writing1244—
EQ-Bench Creative Writing752—
LMArena Multi-Turn1257—

Frequently asked questions

Is GPT-4 better than Step 3.7 Flash?

Step 3.7 Flash is the stronger model overall, scoring 37.3 to 29.1 on the Noometry Index.

Which is cheaper, GPT-4 or Step 3.7 Flash?

Step 3.7 Flash is cheaper. It lists at $0.18 per million input tokens and $1.11 per million output tokens; GPT-4 lists at $30 and $60.

Is GPT-4 or Step 3.7 Flash better for coding?

Step 3.7 Flash scores higher on coding benchmarks: 40.0 versus 31.6 in the Noometry coding category.

Which has the bigger context window?

Step 3.7 Flash does, with 256K tokens against 8K.

How many benchmarks do GPT-4 and Step 3.7 Flash share?

0 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Step 3.7 Flash has 5.

Related comparisons

Go deeper