GPT-4 is the original GPT-4 model launched in March 2023, now deprecated and replaced by GPT-4o.

Is GPT-4 still available?

GPT-4 is deprecated. Migrate all workloads to GPT-4o immediately.

It was groundbreaking at launch, passing the bar exam in the top 10% and achieving human-level performance on many benchmarks.

What should I use instead of GPT-4?

Use GPT-4o, which is 2x faster, 50-75% cheaper, and supports audio.

How much does GPT-4 cost?

GPT-4 pricing is $30.00 per 1M input tokens and $60.00 per 1M output tokens.

LangFast

Beta

GPT-4

Original GPT-4. Kept for legacy compatibility.

Try in Playground

Community Sentiment

Mostly Positive

Based on Reddit reviews

Community Verdict

Best for General Tasks

Based on Reddit reviews

Input Modalities

Text

Output Modalities

Text

Price / 1M tokens

$30$60

InputOutput

Best For

Analytical thinking and tradeoff exploration

Structured problem analysis (with human review)

Nuanced writing, editing, and translation

Long chats needing strong coherence and memory

Avoid For

Simple/repetitive tasks (use cheaper models)

Cost-sensitive, high-volume workloads

Deterministic / strict-format tasks (schemas, pipelines)

High-reliability production flows

8,192 context window

8,192 max output tokens

Sep 1, 2021 knowledge cutoff

Reasoning not supported

Parameters

While OpenAI documents a unified parameter set for the Chat Completions and Responses APIs, each model supports only a limited subset of model-specific parameters and values. This table lists the supported model-specific GPT-4 API parameters and allowed values.

Parameter	Description	Path	Supported Values
Model	Selects the model version for the request	model	gpt-4gpt-4-0613
Message roles	Defines the role of the message in the input	messages[].role	developersystemuserassistant
Max output tokens	Max output tokens the model may generate	max_completion_tokens	16 .. 8,192
Output format	Specifies the output format, including structured JSON schemas	response_format	textjson_object
Temperature	Controls how random or deterministic the output is	temperature	0 .. 2
Top P	Controls how diverse the output tokens are	top_p	0 .. 1
Presence penalty	Encourages the model to introduce new topics	presence_penalty	-2 .. 2
Frequency penalty	Reduces repetition of the same words or phrases	frequency_penalty	-2 .. 2
Reasoning effort	Controls the depth of internal reasoning used by the model	reasoning_effort	Not supported
Reasoning summary	Controls whether the model produces a concise or detailed reasoning summary	reasoning_summary	Not supported
Verbosity	Controls how brief or detailed the generated response is	verbosity	Not supported

Pricing

GPT-4 API pricing is based on token usage for input and output. Prices are listed per 1M tokens, with lower rates for cached input. Tool-specific features may add per-call fees.

Text tokens

Per 1M tokens

Input$30.00

Cached inputN/A

Output$60.00

Example costs (GPT-4)

TASK

APPROX COST

Chat conversation

~$0.03–$0.10

Document analysis

~$0.10–$0.30

Code generation

~$0.05–$0.20

Modalities

What the model can accept and produce

TextInput and output

ImagesNot supported

AudioNot supported

VideoNot supported

Features

Platform-level capabilities

StreamingSupported

Function callingNot supported

Structured outputsNot supported

Fine-tuningSupported

DistillationNot supported

Predicted outputsNot supported

Tools

Tools supported by this model when using the Responses API.

Web searchNot supported

File searchSupported

Image generationNot supported

Code interpreterSupported

MCPSupported

Computer useNot supported

Snapshots

GPT-4 model snapshots ensure stable behavior by locking a specific version. See all available snapshots and aliases below.

gpt-4

↪ gpt-4-0613

gpt-4-0613

gpt-4-0314Deprecated

FAQs

Ship prompts that pass the tests

Don't wait until they break in production

Get started – it's free