LANGUAGE MODEL Mar 2023 OpenAI Last updated: Apr 29, 2026

GPT-4

OpenAI's First Multimodal Frontier Model

OpenAI's successor to GPT-3.5, launched as an API-gated upgrade to ChatGPT Plus subscribers. Demonstrated a step-change in reasoning benchmarks — bar exam, medical licensing, GRE, Olympiad math — and added image input capability (text output only). OpenAI declined to disclose model size or architecture, citing competitive and safety reasons.

Try ChatGPT API Docs ↗

Intelligence

Below avg

Speed

Slow

35 tok/s output

Cost

High

$10.00 in / $30.00 out

Context

Up to 8,192 tokens

How are Intelligence, Speed & Cost bucketed?

Intelligence and Speed buckets are percentile ranks on Artificial Analysis. Cost buckets are fixed dollar thresholds keyed off output-token price ($/M out).

Intelligence

Top 1%≤ 1%
Top 5%≤ 5%
Top 10%≤ 10%
Good≤ 25%
Medium≤ 50%
Below avg> 50%

Speed

Top 1%≥ 345 tok/s
Top 5%≥ 237 tok/s
Top 10%≥ 196 tok/s
Good≥ 146 tok/s
Medium≥ 90 tok/s
Slow< 90 tok/s

Cost

Freeopen weights · self-host
Low< $1 / M out
Moderate$1–5 / M out
High≥ $5 / M out

Official ↗ Artificial Analysis ↗

Why it matters

GPT-4 is what people actually mean when they say "AI" in a business conversation in 2024-25. The reasoning, bar-exam, code-generation, and image-understanding demos from GPT-4's launch are the reference point that every subsequent model is implicitly benchmarked against in casual discourse.

Core Capabilities

Long Documents

Handles entire codebases, books, and multi-doc RAG.

Multimodal

Combines text, vision, and audio in one model.

Generative

Produces images, video, audio, or other media.

Agent Workflows

Built for tool use and autonomous tasks.

Context Window

8k tokens

≈ short doc

4k Chat 聊天

32k Long docs 长文档

128k Books 整本书

400k Multi-doc 多文档

1M Codebase 整个代码库

10M

Availability

API

Available

Product / App

Available

Open Source

Not released

Enterprise

Contact sales

Pricing Model

Pay per token

Input and output billed separately.

Pay-per-token

Capability / Performance

Where this model sits relative to the middle 60% of models in the tree. All scores are 0–10 (higher is better).

Lower 20% Upper 80% This model

Reasoning

AA Intelligence Index · scaled to 10

1.7

5.6

2.0

Coding

SciCode · scaled to 10

1.8

4.3

4.0

Agentic tasks

Terminal-Bench Hard · scaled to 10

0.2

3.6

1.4

Context / memory

Context window size · log-scaled

6.0

9.0

2.0

Cost efficiency

Input price ($/M tokens) · cheaper scores higher

6.2

10.0

1.9

Lower 20% 20th percentile — 20% of models score below this This model Where the current model lands Upper 80% 80th percentile — only 20% of models score above this Percentile boundaries are computed across every model in the tree that reports the underlying benchmark for each capability.

What it feels like

Step-change in reasoning vs GPT-3.5 — top 10% on simulated bar exam vs 3.5's bottom 10%
MMLU 86.4% in English; surpassed prior models in 24 of 26 other languages
First widely-deployed model with image input (text output only) — multimodal era starts here
19 percentage points fewer hallucinations than GPT-3.5 on adversarial factuality tests
Genuinely creative and reliable on nuanced instructions where 3.5 broke
OpenAI declined to publish parameter count or training details — closed-source standard set here