Command Palette

Search for a command to run...

GPT-4o (Nov '24)

OpenAI

Vision ModelProprietary

Context

Release Date
Nov 20, 2024
Knowledge Cutoff
Oct 01, 2023
Window
128k

PricingPer 1M tokens

Input
$2.5
Output
$10
Blended 3:1
$4.375

Capabilities

Speed
159 t/s
Input
Output
Reasoning tokens

Latency

TTFT
0.40 ms
500 token response
3.55 s

Benchmarks

Intelligence
●●○○○
Math
●●○○○
Coding
●●○○○
MMLU Pro
74.8%
GPQA
54.3%
HLE
3.3%
SciCode
33.3%
AIME
15.0%
MATH 500
75.9%
LiveCodeBench
30.9%
HumanEval
93.0%

GPT-4o is OpenAI's latest multimodal model that handles text, code, images, audio, and video in one network. It matches GPT-4 Turbo on English tasks, is stronger in other languages, and offers a 128k context window with ~300 ms speech response at half the price.

Developers get a single, cheaper endpoint for chat, coding, voice, and vision features without juggling multiple models or pipelines. The large context and real-time audio enable fast assistants and rich multimodal analysis in any app.