Saltar al contenido principal

GPT-4.5 (Preview)

OpenAIGPTProprietary

Descripción

GPT-4.5 is OpenAI's most advanced model, offering improved reasoning, coding, and creative capabilities with faster performance and longer context handling than GPT-4. It features enhanced instruction following, reduced hallucinations, and better factual accuracy.

Fecha de lanzamiento
2025-02-27
Parámetros
—
Longitud del contexto
—
Modalidades
image, text

Radar de capacidades

10
general
50
coding
80
reasoning
60
scienceest.
60
agents
70
multimodal

Science se estima a partir de las puntuaciones científicas de LLM Stats o del razonamiento cuando no hay benchmarks científicos dedicados.

Rankings

Dominio#PosiciónPuntuaciónFuente
Ranking general637
12.0
AA
Ranking multimodal67
54.0
LS

Puntuaciones de benchmarks (LLM Stats)

(LLM Stats (zeroeval))

Chat

IFEvalGoogle Research (2023)88.2%Aut.
Multi-IF70.8%Aut.
TAU-bench Retail68.4%Aut.
Multi-Challenge43.8%Aut.

Factuality

SimpleQA62.5%Aut.

General

MMLU90.8%Aut.
Internal API instruction following (hard)54.0%Aut.
Aider-Polyglot Edit44.9%Aut.

Language

MMMLU85.1%Aut.
COLLIE72.3%Aut.

Long Context

ComplexFuncBench63.0%Aut.
OpenAI-MRCR: 2 needle 128k38.5%Aut.

Math

GSM8k97.0%Aut.
MathVista72.3%Aut.
AIME 202436.7%Aut.

Multimodal

MMMU75.2%Aut.

Reasoning

CharXiv-D90.0%Aut.
HumanEvalOpenAI (2021)88.0%Aut.
Graphwalks parents <128k72.6%Aut.
Graphwalks BFS <128k72.3%Aut.
GPQANYU + Cohere + Anthropic (2023)69.5%Aut.
CharXiv-R55.4%Aut.
TAU-bench Airline50.0%Aut.
SWE-Bench Verified38.0%Aut.
SWE-Lancer37.3%Aut.
SWE-Lancer (IC-Diamond subset)17.4%Aut.

Índices de evaluación AA

(Artificial Analysis)
Intelligence Index(Artificial Analysis)
9.6

Puntuaciones por categoría LLM Stats

(LLM Stats (zeroeval))
Legal
90
Finance
90
Instruction Following
80
Language
80
Math
80
Healthcare
80
Chat
70
Multimodal
70
Physics
70
Spatial Reasoning
70
Structured Output
70
Biology
70
Chemistry
70
Vision
70
Writing
70
Reasoning
60
Factuality
60
General
60
Communication
60
Tool Calling
60
Long Context
50
Code
50
Frontend Development
40

Precios

Precio de entradaGratis
Precio de salidaGratis
Precio mixto (3:1)Gratis

Velocidad

Tokens/seg0.0
Retraso del primer token0.00s
Tiempo hasta la respuesta0.00s

Ranking de Precios por Proveedor

No hay datos de proveedores disponibles

Fuentes externas