Closed SourceOpenAIReleased 2024-05

GPT-4o

OpenAI's multimodal workhorse - fast, affordable, and widely integrated

Context window

128K

Input / 1M tokens

$2.50

Output / 1M tokens

$10.00

Provider

OpenAI

GPT-4o ("omni") combines text, image, and audio understanding in a single model endpoint. It matches GPT-4 Turbo quality at 50% of the cost and significantly faster latency. The default model powering ChatGPT, it's the most widely deployed frontier model in the ecosystem.

Strengths

  • Native multimodal input (text, image, audio)
  • Fast response times at this capability level
  • Strong on structured data and JSON output
  • Best ecosystem support across SDKs and tools
  • Real-time audio capabilities

Best for developers who...

Multimodal appsHigh-volume production useChatbots and assistantsStructured data extraction

Benchmarks

BenchmarkScoreNotes
MMLU88.7%Matches GPT-4 Turbo
HumanEval90.2%Strong Python coding
GPQA53.6%Graduate-level science

Source: OpenAI GPT-4o launch post

Tools powered by GPT-4o

Compare GPT-4o with

All model comparisons →

Learn the concepts