Back to all models
G

GPT-4.5

OpenAIReleased 2025-02-27
Official Page

OpenAI's largest and most capable general-purpose model with deep knowledge and reasoning. Delivers state-of-the-art on many benchmarks.

Overall Score
86%
Cost Efficiency
44
Context Window
128K

Strengths

  • ✓Highest knowledge scores
  • ✓Strong overall performance
  • ✓Better instruction following

Weaknesses

  • ✗Very expensive
  • ✗Not a reasoning-specialized model
Architecture
Transformer (MoE)
Parameters
Unknown
Input
$75/M
Output
$150/M

Benchmark Performance

BenchmarkDescriptionScoreCategoryBar
MMLUMassive Multitask Language Understanding91.5%knowledge
HumanEvalCode generation accuracy92.1%coding
MATHMathematical reasoning94.5%math
GSM8KGrade school math95.8%math
GPQAGraduate-level science Q&A73.3%reasoning
HellaSwagCommon sense reasoning89.5%reasoning
SWE-benchSoftware engineering problems52.4%coding
ARC-CAI2 Reasoning Challenge97%reasoning
AI
ModelBench

The open benchmark platform for comparing the latest AI models and their capabilities.

Platform

  • Compare Models
  • All Models
  • Methodology

AI Labs

  • OpenAI
  • Anthropic
  • Google DeepMind
  • Meta AI

Resources

  • Documentation
  • Benchmark Guide
  • API Reference
Benchmarks are sourced from official model cards and independent evaluations. Data is for informational purposes only.
© 2026 ModelBench. Built for the AI community.