OpenAI logo

GPT-4.5

OpenAIProprietaryVerified

A research preview of OpenAI's largest and most capable GPT model to date. GPT-4.5 represents a significant advancement in scaling unsupervised learning, delivering broader world knowledge, dramatically reduced hallucinations (37.1% vs 61.8% for GPT-4o), and enhanced emotional intelligence for more natural human collaboration. Excels at creative tasks, writing assistance, and complex problem-solving with improved factual accuracy and nuanced understanding of user intent.

2025-02-27
Unsupervised learning at scale with transformer architecture
Proprietary

Specifications

Architecture
Unsupervised learning at scale with transformer architecture
License
Proprietary
Context Window
128,000 tokens
Max Output
16,384 tokens
Training Data Cutoff
October 2023
Type
multimodal
Modalities
textimage

Benchmark Scores

GPQA71.4

Graduate-level Problems in Quantitative Analysis (GPQA) evaluates advanced reasoning on graduate-lev...

AIME-202436.7

American Invitational Mathematics Examination (AIME) 2024 problems....

MMLU85.1

Massive Multitask Language Understanding (MMLU) tests knowledge across 57 subjects including mathema...

MMMU74.4

Massive Multi-discipline Multimodal Understanding (MMMU) evaluates multimodal understanding across 3...

SWE-Lancer186,125

Evaluates models on real-world freelance software engineering tasks....

SWE-bench38

Software Engineering Benchmark (SWE-bench) evaluates models on real-world software engineering tasks...

See all benchmarks

Advanced Specifications

Model Family
GPT
API Access
Available
Chat Interface
Available
Multilingual Support
Yes

Capabilities & Limitations

Capabilities
Broad world knowledge with fewer hallucinationsImproved emotional intelligence and nuanced understandingExcels at creative writing and design tasksStrong at coding assistance and complex task automationSupports function calling and image inputs (tool use)
Known Limitations
Lacks chain-of-thought reasoning for complex logic/mathsSlower response times due to its large sizeCannot produce audio or video outputs (text-only replies)Still may hallucinate or err on factual queries
Notable Use Cases
Creative writing assistance and content brainstormingPersonalized learning, coaching, and communication helpMulti-step coding workflows and task automation
Function Calling Support
Yes
Tool Use Support
Yes

Related Models