Noometry

Model comparison

GPT-4 Turbo vs Hy3

Hy3 is the stronger model overall, scoring 44.2 to 30.5 on the Noometry Index.

Last verified . 17 shared benchmarks.

GPT-4 Turbo OpenAI

30.5

Rank #292 Confirmed

Hy3 Tencent

44.2

Rank #79 Confirmed

Summary

  • They share 17 benchmarks with published results for both. GPT-4 Turbo scores higher in 0 categories and Hy3 in 8 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Hy3 leads 40.1 to 9.0.
  • Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
  • Hy3 accepts more context: 262K tokens versus 128K.
  • Hy3 has downloadable open weights; the other is API-only.

Side by side

GPT-4 TurboHy3
ProviderOpenAITencent
Noometry Index30.544.2
Released2023-11-062026-07-06
WeightsProprietaryOpen
Context window128K262K
Max output4K128K
Input $ / M tokens$10$0.0825
Output $ / M tokens$30$0.33
Results tracked3619

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

Category by category

Coding Hy3 leads

GPT-4 Turbo: 33.8 (#249), Hy3: 46.8 (#63)

BenchmarkGPT-4 TurboHy3
LMArena Coding12681464
LMArena WebDev—1508
WeirdML18%—
BigCodeBench Instruct48.2%—
BigCodeBench Complete58.2%—
HumanEval+86.6%—
MBPP+73.3%—

Agentic & Tool Use Not comparable

GPT-4 Turbo: —, Hy3: —

BenchmarkGPT-4 TurboHy3
METR Time Horizons36.7%—

Reasoning Hy3 leads

GPT-4 Turbo: 15.3 (#317), Hy3: 26.1 (#136)

BenchmarkGPT-4 TurboHy3
LMArena Hard Prompts12511447
SimpleBench25.1%—
NYT Connections (extended)—41.2%
Chess Puzzles6%—
DTBench61.6%—
LMCA9.8%—
Epoch Capabilities Index127.25—
ForecastBench59.4—

Math Hy3 leads

GPT-4 Turbo: 9.0 (#322), Hy3: 40.1 (#93)

BenchmarkGPT-4 TurboHy3
LMArena Math12721475
FrontierMath (Tiers 1-3)0.7%—
OTIS Mock AIME 2024-20256.7%—
MATH Level 546.7%—

Knowledge Hy3 leads

GPT-4 Turbo: 24.3 (#268), Hy3: 40.8 (#114)

BenchmarkGPT-4 TurboHy3
LMArena Expert12231460
GPQA Diamond46.6%—
Confabulations28.4%—
MMLU81.3%—

Multimodal Not comparable

GPT-4 Turbo: 30.6 (#110), Hy3: —

BenchmarkGPT-4 TurboHy3
LMArena Vision1090—

Multilingual Hy3 leads

GPT-4 Turbo: 40.5 (#216), Hy3: 53.5 (#65)

BenchmarkGPT-4 TurboHy3
LMArena Non-English12451426
LMArena Chinese12421493
LMArena French12761461
LMArena German12591439
LMArena Japanese11941392
LMArena Korean11871395
LMArena Russian12591432
LMArena Spanish12601456

Instruction Following Hy3 leads

GPT-4 Turbo: 65.8 (#216), Hy3: 75.1 (#70)

BenchmarkGPT-4 TurboHy3
LMArena Instruction Following12491426

Long Context Hy3 leads

GPT-4 Turbo: 38.0 (#206), Hy3: 44.1 (#75)

BenchmarkGPT-4 TurboHy3
LMArena Longer Query12541442

Writing & Preference Hy3 leads

GPT-4 Turbo: 47.7 (#206), Hy3: 62.2 (#81)

BenchmarkGPT-4 TurboHy3
LMArena Text12721439
LMArena Creative Writing12691402
LMArena Multi-Turn12671436

Frequently asked questions

Is GPT-4 Turbo better than Hy3?

Hy3 is the stronger model overall, scoring 44.2 to 30.5 on the Noometry Index.

Which is cheaper, GPT-4 Turbo or Hy3?

Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; GPT-4 Turbo lists at $10 and $30.

Is GPT-4 Turbo or Hy3 better for coding?

Hy3 scores higher on coding benchmarks: 46.8 versus 33.8 in the Noometry coding category.

Which has the bigger context window?

Hy3 does, with 262K tokens against 128K.

How many benchmarks do GPT-4 Turbo and Hy3 share?

17 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and Hy3 has 19.

Related comparisons

Go deeper