Noometry

Model comparison

Claude 3 Haiku vs GPT-4.1

GPT-4.1 is the stronger model overall, scoring 35.9 to 25.9 on the Noometry Index.

Last verified . 29 shared benchmarks.

Claude 3 Haiku Anthropic

25.9

Rank #340 Confirmed

GPT-4.1 OpenAI

35.9

Rank #219 Confirmed

Summary

  • They share 29 benchmarks with published results for both. Claude 3 Haiku scores higher in 1 category and GPT-4.1 in 8 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where GPT-4.1 leads 57.6 to 29.7.
  • The biggest single-benchmark swing is MATH Level 5: 14.9% for Claude 3 Haiku and 83% for GPT-4.1.

Side by side

Claude 3 HaikuGPT-4.1
ProviderAnthropicOpenAI
Noometry Index25.935.9
Released2024-03-072025-04-14
WeightsProprietaryProprietary
Context window—1.05M
Max output—33K
Input $ / M tokens—$2
Output $ / M tokens—$8
Results tracked3752

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

Category by category

Coding GPT-4.1 leads

Claude 3 Haiku: 26.4 (#325), GPT-4.1: 34.4 (#238)

BenchmarkClaude 3 HaikuGPT-4.1
WeirdML9.8%39%
LMArena Coding11991391
CadEval12%42%
SWE-bench Verified—48.5%
SWE-bench Verified (bash only)—39.6%
Aider Polyglot—52.4%
BigCodeBench Instruct39.4%—
BigCodeBench Complete50.1%—
ALE-Bench—558.1
HumanEval+68.9%—
MBPP+68.8%—

Agentic & Tool Use Not comparable

Claude 3 Haiku: —, GPT-4.1: 34.7 (#43)

BenchmarkClaude 3 HaikuGPT-4.1
Berkeley Function Calling Leaderboard—54%

Reasoning Claude 3 Haiku leads

Claude 3 Haiku: 16.3 (#307), GPT-4.1: 11.7 (#339)

BenchmarkClaude 3 HaikuGPT-4.1
Kagi LLM Benchmark34.2%52.3%
LMArena Hard Prompts11741384
DTBench50.1%68.3%
LMCA8.8%25.6%
Epoch Capabilities Index118.35136.78
ForecastBench53.261.5
ARC-AGI-2—0.4%
SimpleBench—27%
ARC-AGI-1—5.5%
Chess Puzzles—6%
EnigmaEval—2.2%
WinoGrande74.2%—

Math GPT-4.1 leads

Claude 3 Haiku: 9.8 (#319), GPT-4.1: 22.3 (#280)

BenchmarkClaude 3 HaikuGPT-4.1
OTIS Mock AIME 2024-20251.8%38.3%
LMArena Math11881370
MATH Level 514.9%83%
FrontierMath (Tiers 1-3)—6%
Omni-MATH—47.1%
FrontierMath (Feb 2025 set)—5.5%
FrontierMath Tier 4 (v1)—0%

Knowledge GPT-4.1 leads

Claude 3 Haiku: 17.3 (#285), GPT-4.1: 37.1 (#160)

BenchmarkClaude 3 HaikuGPT-4.1
GPQA Diamond36.3%66.9%
LMArena Expert11481364
Humanity's Last Exam—5.4%
SimpleQA Verified—31.1%
MMLU-Pro—81.1%
Confabulations34.2%—
Vectara Hallucination Rate—5.6%
GPQA (HELM)—65.9%
MMLU73.8%—

Multimodal GPT-4.1 leads

Claude 3 Haiku: 23.6 (#128), GPT-4.1: 38.2 (#67)

BenchmarkClaude 3 HaikuGPT-4.1
LMArena Vision9501211
GeoBench—72%
ScienceQA72%—

Multilingual GPT-4.1 leads

Claude 3 Haiku: 36.0 (#243), GPT-4.1: 49.4 (#133)

BenchmarkClaude 3 HaikuGPT-4.1
LMArena Non-English11781370
LMArena Chinese11551382
LMArena French11951382
LMArena German11741381
LMArena Japanese11021319
LMArena Korean11091339
LMArena Russian12041377
LMArena Spanish11661376

Instruction Following GPT-4.1 leads

Claude 3 Haiku: 61.3 (#247), GPT-4.1: 71.3 (#153)

BenchmarkClaude 3 HaikuGPT-4.1
LMArena Instruction Following11731367
IFEval—83.8%

Long Context GPT-4.1 leads

Claude 3 Haiku: 36.1 (#237), GPT-4.1: 40.0 (#163)

BenchmarkClaude 3 HaikuGPT-4.1
LMArena Longer Query11901385
Fiction.LiveBench—63.9%

Writing & Preference GPT-4.1 leads

Claude 3 Haiku: 29.7 (#291), GPT-4.1: 57.6 (#125)

BenchmarkClaude 3 HaikuGPT-4.1
LMArena Text11951383
LMArena Creative Writing11571363
EQ-Bench Creative Writing7171420
LMArena Multi-Turn11901398
WildBench—85.4%

Frequently asked questions

Is Claude 3 Haiku better than GPT-4.1?

GPT-4.1 is the stronger model overall, scoring 35.9 to 25.9 on the Noometry Index.

Is Claude 3 Haiku or GPT-4.1 better for coding?

GPT-4.1 scores higher on coding benchmarks: 34.4 versus 26.4 in the Noometry coding category.

How many benchmarks do Claude 3 Haiku and GPT-4.1 share?

29 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and GPT-4.1 has 52.

Related comparisons

Go deeper