Model comparison

Amazon Nova Lite vs GPT-4 Turbo

Amazon Nova Lite is the stronger model overall, scoring 31.9 to 30.5 on the Noometry Index.

Last verified . 19 shared benchmarks.

Amazon Nova Lite Amazon

31.9

Rank #265 Confirmed

GPT-4 Turbo OpenAI

30.5

Rank #292 Confirmed

Summary

  • They share 19 benchmarks with published results for both. Amazon Nova Lite scores higher in 3 categories and GPT-4 Turbo in 6 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Amazon Nova Lite leads 27.8 to 9.0.
  • Amazon Nova Lite is cheaper at $0.06 / $0.24 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
  • Amazon Nova Lite accepts more context: 300K tokens versus 128K.

Side by side

Amazon Nova Lite and GPT-4 Turbo specifications
Amazon Nova LiteGPT-4 Turbo
ProviderAmazonOpenAI
Noometry Index31.930.5
Released2024-12-032023-11-06
WeightsProprietaryProprietary
Context window300K128K
Max output10K4K
Input $ / M tokens$0.06$10
Output $ / M tokens$0.24$30
Results tracked3436

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

Category by category

Coding GPT-4 Turbo leads

Amazon Nova Lite: 32.5 (#270), GPT-4 Turbo: 33.8 (#249)

Coding benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Coding12391268
WeirdML—18%
BigCodeBench Instruct—48.2%
LiveBench Coding27.5%—
BigCodeBench Complete—58.2%
ALE-Bench236.25—
HumanEval+—86.6%
MBPP+—73.3%

Agentic & Tool Use Not comparable

Amazon Nova Lite: —, GPT-4 Turbo: —

Agentic & Tool Use benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
METR Time Horizons—36.7%

Reasoning Amazon Nova Lite leads

Amazon Nova Lite: 19.3 (#260), GPT-4 Turbo: 15.3 (#317)

Reasoning benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Hard Prompts12201251
SimpleBench—25.1%
Chess Puzzles—6%
LiveBench Reasoning36.7%—
DTBench—61.6%
LiveBench Data Analysis37.2%—
LMCA—9.8%
Epoch Capabilities Index—127.25
ForecastBench—59.4
LiveBench36.4%—

Math Amazon Nova Lite leads

Amazon Nova Lite: 27.8 (#247), GPT-4 Turbo: 9.0 (#322)

Math benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Math12271272
FrontierMath (Tiers 1-3)—0.7%
OTIS Mock AIME 2024-2025—6.7%
Omni-MATH23.3%—
LiveBench Math36.7%—
MATH Level 5—46.7%

Knowledge Amazon Nova Lite leads

Amazon Nova Lite: 26.3 (#259), GPT-4 Turbo: 24.3 (#268)

Knowledge benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Expert12011223
MMLU77%81.3%
GPQA Diamond—46.6%
Humanity's Last Exam3.6%—
MMLU-Pro60%—
Confabulations—28.4%
Vectara Hallucination Rate6.1%—
GPQA (HELM)39.7%—

Multimodal GPT-4 Turbo leads

Amazon Nova Lite: 25.5 (#123), GPT-4 Turbo: 30.6 (#110)

Multimodal benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Vision9901090

Multilingual GPT-4 Turbo leads

Amazon Nova Lite: 38.0 (#232), GPT-4 Turbo: 40.5 (#216)

Multilingual benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Non-English12081245
LMArena Chinese12251242
LMArena French12371276
LMArena German12291259
LMArena Japanese11531194
LMArena Korean11531187
LMArena Russian12161259
LMArena Spanish12261260

Instruction Following GPT-4 Turbo leads

Amazon Nova Lite: 59.3 (#255), GPT-4 Turbo: 65.8 (#216)

Instruction Following benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Instruction Following12051249
LiveBench Instruction Following54.1%—
IFEval77.6%—

Long Context Too close to call

Amazon Nova Lite: 37.4 (#217), GPT-4 Turbo: 38.0 (#206)

Long Context benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Longer Query12341254

Writing & Preference GPT-4 Turbo leads

Amazon Nova Lite: 42.1 (#237), GPT-4 Turbo: 47.7 (#206)

Writing & Preference benchmarks
BenchmarkAmazon Nova LiteGPT-4 Turbo
LMArena Text12291272
LMArena Creative Writing11971269
LMArena Multi-Turn11991267
WildBench75%—
LiveBench Language25.9%—

Frequently asked questions

Is Amazon Nova Lite better than GPT-4 Turbo?

Amazon Nova Lite is the stronger model overall, scoring 31.9 to 30.5 on the Noometry Index.

Which is cheaper, Amazon Nova Lite or GPT-4 Turbo?

Amazon Nova Lite is cheaper. It lists at $0.06 per million input tokens and $0.24 per million output tokens; GPT-4 Turbo lists at $10 and $30.

Is Amazon Nova Lite or GPT-4 Turbo better for coding?

GPT-4 Turbo scores higher on coding benchmarks: 33.8 versus 32.5 in the Noometry coding category.

Which has the bigger context window?

Amazon Nova Lite does, with 300K tokens against 128K.

How many benchmarks do Amazon Nova Lite and GPT-4 Turbo share?

19 benchmarks have published results for both models. Amazon Nova Lite has 34 scored results on Noometry and GPT-4 Turbo has 36.

Related comparisons

Go deeper