GPT-5 Nano is the ultra-cheap variant of GPT-5, optimized for classification, extraction, routing, and other simple tasks at minimal cost.

How good is GPT-5 Nano?

GPT-5 Nano offers basic capabilities at the lowest cost, though it may struggle with complex tasks, multilingual content, and formatted outputs.

What can GPT-5 Nano do?

It excels at ultra-low-cost classification, simple extraction, routing/triage, and high-volume English text processing.

What are GPT-5 Nano's limitations?

Poor caching reliability (~10% hit rate), limited multilingual support, and inconsistent markdown formatting for code/math.

How much does GPT-5 Nano cost?

GPT-5 Nano pricing is $0.05 per 1M input tokens and $0.40 per 1M output tokens. Cached input tokens cost $0.005 per 1M.

LangFast

Beta

GPT-5 Nano

Ultra-cheap GPT-5 for classification, extraction, and routing.

Try in Playground

Community Sentiment

Mixed

Based on Reddit reviews

Community Verdict

Best for Ultra-Budget

Based on Reddit reviews

Input Modalities

Text, Images

Output Modalities

Text

Price / 1M tokens

$0.05$0.4

InputOutput

Best For

Ultra-low-cost use where budget dominates

Simple English-only text tasks

High-volume, low-complexity workloads

Use cases tolerant of minor inconsistencies

Basic classification and simple extraction

Avoid For

Workloads relying on caching (low cache hit rates)

Code/math outputs that must be in markdown

Multilingual apps (weak non-English)

Tasks needing consistent, well-formatted output

Quality-critical applications

Complex reasoning or nuanced understanding

128,000 context window

65,536 max output tokens

Jan 1, 2024 knowledge cutoff

Reasoning 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-5 Nano API parameters and allowed values.

Parameter	Description	Path	Supported Values
Model	Selects the model version for the request	model	gpt-5-nano
Message roles	Defines the role of the message in the input	messages[].role	developersystemuserassistant
Reasoning effort	Controls the depth of internal reasoning used by the model	reasoning_effort	minimallowmediumhigh
Reasoning summary	Controls whether the model produces a concise or detailed reasoning summary	reasoning_summary	detailedconciseautonull
Max output tokens	Max output tokens the model may generate	max_completion_tokens	16 .. 65,536
Verbosity	Controls how brief or detailed the generated response is	verbosity	lowmediumhigh
Output format	Specifies the output format, including structured JSON schemas	response_format	textjson_objectjson_schema
Temperature	Controls how random or deterministic the output is	temperature	Not supported
Top P	Controls how diverse the output tokens are	top_p	Not supported
Presence penalty	Encourages the model to introduce new topics	presence_penalty	Not supported
Frequency penalty	Reduces repetition of the same words or phrases	frequency_penalty	Not supported

Pricing

GPT-5 Nano 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$0.05

Cached input$0.01

Output$0.40

Example costs (GPT-5 Nano)

TASK

APPROX COST

Simple classification (100 tokens)

~$0.00001

Text extraction batch (10k items)

~$0.05

Routing/triage (1M requests)

~$0.50

Modalities

What the model can accept and produce

TextInput and output

ImagesInput only

AudioNot supported

VideoNot supported

Features

Platform-level capabilities

StreamingSupported

Function callingSupported

Structured outputsSupported

Fine-tuningNot supported

DistillationNot supported

Predicted outputsNot supported

Tools

Tools supported by this model when using the Responses API.

Web searchSupported

File searchSupported

Image generationSupported

Code interpreterSupported

MCPSupported

Computer useNot supported

Snapshots

GPT-5 Nano model snapshots ensure stable behavior by locking a specific version. See all available snapshots and aliases below.

gpt-5-nano

↪ gpt-5-nano-2025-08-07

gpt-5-nano-2025-08-07

FAQs

Ship prompts that pass the tests

Don't wait until they break in production

Get started – it's free