Noometry

Model comparison

Claude 3 Haiku vs GPT-6 Sol

GPT-6 Sol is the stronger model overall, scoring 61.8 to 25.9 on the Noometry Index.

Last verified . 24 shared benchmarks.

Claude 3 Haiku Anthropic

25.9

Rank #340 Confirmed

GPT-6 Sol OpenAI

61.8

Rank #12 Confirmed

Summary

  • They share 24 benchmarks with published results for both. Claude 3 Haiku scores higher in 0 categories and GPT-6 Sol in 9 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in math, where GPT-6 Sol leads 87.2 to 9.8.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.8% for Claude 3 Haiku and 100% for GPT-6 Sol.

Side by side

Claude 3 HaikuGPT-6 Sol
ProviderAnthropicOpenAI
Noometry Index25.961.8
Released2024-03-072026-09-22
WeightsProprietaryProprietary
Context window—1.05M
Max output—128K
Input $ / M tokens—$2
Output $ / M tokens—$10
Results tracked3745

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

Category by category

Coding GPT-6 Sol leads

Claude 3 Haiku: 26.4 (#325), GPT-6 Sol: 60.1 (#11)

BenchmarkClaude 3 HaikuGPT-6 Sol
LMArena Coding11991447
DeepSWE—68.8%
FrontierCode—49.3%
LMArena WebDev—1688
SciCode—57.6%
WeirdML9.8%—
BigCodeBench Instruct39.4%—
BigCodeBench Complete50.1%—
CadEval12%—
ALE-Bench—2,462
HumanEval+68.9%—
MBPP+68.8%—

Agentic & Tool Use Not comparable

Claude 3 Haiku: —, GPT-6 Sol: 37.2 (#36)

BenchmarkClaude 3 HaikuGPT-6 Sol
APEX-Agents—54.3%
GDP.pdf—26.4%
Vending-Bench 2—14,428

Reasoning GPT-6 Sol leads

Claude 3 Haiku: 16.3 (#307), GPT-6 Sol: 74.0 (#9)

BenchmarkClaude 3 HaikuGPT-6 Sol
LMArena Hard Prompts11741418
DTBench50.1%97.3%
LMCA8.8%59.1%
Epoch Capabilities Index118.35162.72
ARC-AGI-2—89.6%
Kagi LLM Benchmark34.2%—
NYT Connections (extended)—90.1%
ARC-AGI-1—95.5%
CritPt—30.9%
EBR-Bench—53.3%
Mystery Game Puzzles—56%
ForecastBench53.2—
WinoGrande74.2%—

Math GPT-6 Sol leads

Claude 3 Haiku: 9.8 (#319), GPT-6 Sol: 87.2 (#7)

BenchmarkClaude 3 HaikuGPT-6 Sol
OTIS Mock AIME 2024-20251.8%100%
LMArena Math11881402
FrontierMath (Tiers 1-3)—89.8%
FrontierMath Tier 4—90%
ProofBench—83%
MATH Level 514.9%—

Knowledge GPT-6 Sol leads

Claude 3 Haiku: 17.3 (#285), GPT-6 Sol: 64.8 (#15)

BenchmarkClaude 3 HaikuGPT-6 Sol
GPQA Diamond36.3%94.3%
LMArena Expert11481439
SimpleQA Verified—60.7%
Confabulations34.2%—
Vectara Hallucination Rate—6.5%
MMLU73.8%—

Multimodal GPT-6 Sol leads

Claude 3 Haiku: 23.6 (#128), GPT-6 Sol: 47.6 (#10)

BenchmarkClaude 3 HaikuGPT-6 Sol
LMArena Vision9501245
Blueprint-Bench 2—36.9%
Furniture Assembly—58.3%
ScienceQA72%—

Multilingual GPT-6 Sol leads

Claude 3 Haiku: 36.0 (#243), GPT-6 Sol: 50.5 (#118)

BenchmarkClaude 3 HaikuGPT-6 Sol
LMArena Non-English11781385
LMArena Chinese11551405
LMArena French11951410
LMArena German11741390
LMArena Japanese11021385
LMArena Korean11091341
LMArena Russian12041401
LMArena Spanish11661384

Instruction Following GPT-6 Sol leads

Claude 3 Haiku: 61.3 (#247), GPT-6 Sol: 74.5 (#94)

BenchmarkClaude 3 HaikuGPT-6 Sol
LMArena Instruction Following11731412

Long Context GPT-6 Sol leads

Claude 3 Haiku: 36.1 (#237), GPT-6 Sol: 43.1 (#108)

BenchmarkClaude 3 HaikuGPT-6 Sol
LMArena Longer Query11901411

Writing & Preference GPT-6 Sol leads

Claude 3 Haiku: 29.7 (#291), GPT-6 Sol: 71.9 (#18)

BenchmarkClaude 3 HaikuGPT-6 Sol
LMArena Text11951395
LMArena Creative Writing11571378
EQ-Bench Creative Writing7172125
LMArena Multi-Turn11901412

Frequently asked questions

Is Claude 3 Haiku better than GPT-6 Sol?

GPT-6 Sol is the stronger model overall, scoring 61.8 to 25.9 on the Noometry Index.

Is Claude 3 Haiku or GPT-6 Sol better for coding?

GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 26.4 in the Noometry coding category.

How many benchmarks do Claude 3 Haiku and GPT-6 Sol share?

24 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and GPT-6 Sol has 45.

Related comparisons

Go deeper