Noometry

Model comparison

Claude 3 Sonnet vs GPT-4

Claude 3 Sonnet and GPT-4 score almost the same on the Noometry Index (29.0 vs 29.1), so choose on price, context window or the category you care about most.

Last verified . 28 shared benchmarks.

Claude 3 Sonnet Anthropic

29.0

Rank #319 Confirmed

GPT-4 OpenAI

29.1

Rank #316 Confirmed

Summary

  • They share 28 benchmarks with published results for both. Claude 3 Sonnet scores higher in 3 categories and GPT-4 in 5 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where Claude 3 Sonnet leads 42.1 to 34.9.
  • The biggest single-benchmark swing is DTBench: 53.6% for Claude 3 Sonnet and 62.7% for GPT-4.

Side by side

Claude 3 SonnetGPT-4
ProviderAnthropicOpenAI
Noometry Index29.029.1
Released2024-02-292023-03-14
WeightsProprietaryProprietary
Context window—8K
Max output—8K
Input $ / M tokens—$30
Output $ / M tokens—$60
Results tracked3038

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

Category by category

Coding GPT-4 leads

Claude 3 Sonnet: 29.6 (#302), GPT-4: 31.6 (#283)

BenchmarkClaude 3 SonnetGPT-4
WeirdML10.2%12.4%
BigCodeBench Instruct42.7%46%
LMArena Coding12231254
BigCodeBench Complete53.8%57.2%
HumanEval+64%79.3%
MBPP+69.3%—

Agentic & Tool Use Not comparable

Claude 3 Sonnet: —, GPT-4: —

BenchmarkClaude 3 SonnetGPT-4
METR Time Horizons—36.1%

Reasoning Claude 3 Sonnet leads

Claude 3 Sonnet: 20.5 (#237), GPT-4: 17.8 (#289)

BenchmarkClaude 3 SonnetGPT-4
LMArena Hard Prompts11971241
DTBench53.6%62.7%
Epoch Capabilities Index120.7125.89
WinoGrande75.1%87.5%
Chess Puzzles—4%
Mystery Game Puzzles—12%
LMCA—17.1%
BIG-Bench Hard—75.1%
ForecastBench—57.8
HellaSwag—95.3%

Math Too close to call

Claude 3 Sonnet: 10.7 (#310), GPT-4: 10.8 (#309)

BenchmarkClaude 3 SonnetGPT-4
OTIS Mock AIME 2024-20252.5%1.1%
LMArena Math12131269
MATH Level 518.2%23%
GSM8K—92%

Knowledge Claude 3 Sonnet leads

Claude 3 Sonnet: 21.1 (#276), GPT-4: 18.4 (#282)

BenchmarkClaude 3 SonnetGPT-4
GPQA Diamond40.6%35.7%
LMArena Expert11731211
MMLU75.9%86.4%
TriviaQA—84.8%

Multimodal Not comparable

Claude 3 Sonnet: 25.2 (#125), GPT-4: —

BenchmarkClaude 3 SonnetGPT-4
LMArena Vision984—

Multilingual GPT-4 leads

Claude 3 Sonnet: 37.8 (#234), GPT-4: 40.6 (#215)

BenchmarkClaude 3 SonnetGPT-4
LMArena Non-English12051246
LMArena Chinese11891242
LMArena French12291283
LMArena German12041251
LMArena Japanese11311209
LMArena Korean11281184
LMArena Russian12271251
LMArena Spanish12041261

Instruction Following GPT-4 leads

Claude 3 Sonnet: 62.8 (#235), GPT-4: 65.3 (#222)

BenchmarkClaude 3 SonnetGPT-4
LMArena Instruction Following11991241

Long Context GPT-4 leads

Claude 3 Sonnet: 36.7 (#228), GPT-4: 37.7 (#212)

BenchmarkClaude 3 SonnetGPT-4
LMArena Longer Query12111244

Writing & Preference Claude 3 Sonnet leads

Claude 3 Sonnet: 42.1 (#238), GPT-4: 34.9 (#268)

BenchmarkClaude 3 SonnetGPT-4
LMArena Text12181263
LMArena Creative Writing11861244
LMArena Multi-Turn12271257
EQ-Bench Creative Writing—752

Frequently asked questions

Is Claude 3 Sonnet better than GPT-4?

Claude 3 Sonnet and GPT-4 score almost the same on the Noometry Index (29.0 vs 29.1), so choose on price, context window or the category you care about most.

Is Claude 3 Sonnet or GPT-4 better for coding?

GPT-4 scores higher on coding benchmarks: 31.6 versus 29.6 in the Noometry coding category.

How many benchmarks do Claude 3 Sonnet and GPT-4 share?

28 benchmarks have published results for both models. Claude 3 Sonnet has 30 scored results on Noometry and GPT-4 has 38.

Related comparisons

Go deeper