GPT-4.1

OpenAI · openai/gpt-4-1

ProviderOpenAI
Familygpt
Typevlm
Statusactive
Released2025-04-14
Updated2025-04-14

Reported benchmark scores

These scores carry one collection date for the whole card (2026-04) and a source list rather than a source per score, so they cannot be attributed individually. They are shown as reported and are not verified evidence. Benchmark pages carry the reviewed, dated evidence where it exists.
BenchmarkCatalogue standingScore
AI2D (AI2 Diagrams)unassessed95.2
Aider Polyglot Benchmarkunassessed68.5
AlpacaEvalunassessed52.1
Arena Elo — Codingunassessed1360.0
Arena Elo — Hard Promptsunassessed1430.8
Arena Elo — Mathunassessed1320.0
Arena Elo — Overall (Text)unassessed1340.0
Arena Elo — Style Controlunassessed1413.0
Arena Elo — Visionunassessed1213.5
Artificial Analysis Intelligence Indexunassessed84.0
Artificial Analysis Output Speedunassessed76.0
BBQ (Bias Benchmark for QA)unassessed86.5
ChartQAunassessed88.5
DocVQAunassessed94.1
GPQA Diamondunassessed69.5
HELM Safetyunassessed90.5
HumanEvalunassessed90.2
IFEvalunassessed89.5
LiveCodeBenchunassessed48.3
MATH-500unassessed89.4
MathVistaunassessed68.2
MedQAunassessed89.7
MMLU: Astronomyunassessed79.5
MMLU: Biology (subcategory)unassessed87.5
MMLU: Business Ethicsunassessed81.5
MMLU: Chemistry (subcategory)unassessed81.8
MMLU: Clinical Knowledgeunassessed86.2
MMLU: Computer Science (subcategory)unassessed88.2
MMLU: Jurisprudenceunassessed79.8
MMLU: Physics (unresolved key)unassessed84.2
MMLU-Prounassessed80.1
MMLU: Professional Accountingunassessed71.8
MMLU: Professional Lawunassessed77.5
MMMUunassessed72.5
MT-Benchunassessed9.3
MultiPL-Eunassessed85.7
MultiPL-E: C++unassessed85.1
MultiPL-E: C#unassessed85.2
MultiPL-E: Gounassessed82.1
MultiPL-E: Javaunassessed88.5
MultiPL-E: JavaScriptunassessed88.2
MultiPL-E: Juliaunassessed65.8
MultiPL-E: Kotlin (unconfirmed)unassessed78.5
MultiPL-E: Luaunassessed61.5
MultiPL-E: Perlunassessed58.2
MultiPL-E: PHPunassessed82.8
MultiPL-E: Pythonunassessed92.5
MultiPL-E: Runassessed63.5
MultiPL-E: Rubyunassessed73.5
MultiPL-E: Rustunassessed76.2
MultiPL-E: Scalaunassessed67.2
MultiPL-E: Swiftunassessed71.8
MultiPL-E: TypeScriptunassessed87.1
SWE-bench Agentunassessed48.5
SWE-bench Verifiedunassessed54.6
τ-benchunassessed52.8
Terminal-Benchunassessed42.1
ToxiGenunassessed93.5
WildBenchunassessed80.5

Data

This card as JSON · Edit on GitHub