AI/TLDR

GPT-5.4 nano

OpenAI's cheapest GPT-5.4-class model — 400K context, text + image in, built for high-volume classification, extraction, and sub-agents.

Overview

GPT-5.4 nano is the smallest and cheapest model in OpenAI's GPT-5.4 line, released on March 17, 2026 alongside GPT-5.4 mini. OpenAI positions it as "our cheapest GPT-5.4-class model for simple high-volume tasks" — the variant to reach for when speed and cost matter more than deep reasoning. It is recommended for classification, data extraction, ranking, and coding sub-agents that handle simpler supporting work inside a larger system.

GPT-5.4 nano carries the same 400,000-token context window as GPT-5.4 mini and can generate up to 128,000 output tokens. It accepts both text and image input and returns text, and supports reasoning, function calling, structured outputs, and streaming. Its knowledge cutoff is August 31, 2025. OpenAI describes it as a significant upgrade over the previous GPT-5 nano. The model is identified in the API as gpt-5.4-nano, with the dated snapshot gpt-5.4-nano-2026-03-17.

Pricing is $0.20 per million input tokens, $0.02 per million cached input tokens, and $1.25 per million output tokens. GPT-5.4 nano launched in the OpenAI API and is also available through Codex and ChatGPT. As a closed, API-served model it has no published weights or parameter count, so the page below reports only specifications OpenAI has documented.

Released2026-03-17
LicenseProprietary
WeightsAPI only
Context400K
Max output128K
ArchitectureProprietary GPT-5.4-class transformer with reasoning support. OpenAI has not published the parameter count or internal architecture. The model accepts text and image input and returns text, and supports configurable reasoning effort, function calling, structured outputs, and streaming.
Knowledge cutoffAugust 31, 2025
ModalitiesText, Vision
StatusAvailable

Pricing

Input$0.20 / 1M tokens per 1M tokens
Cached input$0.02 / 1M tokens per 1M tokens
Output$1.25 / 1M tokens per 1M tokens

Standard processing tier. GPT-5.4 nano is roughly 4x the price of GPT-5 nano. A Batch API tier is also available.

Pricing source ↗

Strengths

  • Lowest price in the GPT-5.4 family at $0.20 input / $1.25 output per 1M tokens, with cached input at $0.02 per 1M for repeated prompts
  • Large 400K-token context window and 128K max output, matching GPT-5.4 mini despite the lower cost
  • Optimized for low latency and high throughput, making it well suited to high-volume production pipelines
  • Accepts text and image input, and supports function calling, structured outputs, streaming, and configurable reasoning effort
  • A meaningful upgrade over the prior GPT-5 nano while staying in the cheapest tier

Best for

  • Classification and routing of large request volumes
  • Structured data extraction from documents and unstructured text
  • Ranking and scoring tasks where many candidates must be evaluated cheaply
  • Coding sub-agents that handle simpler supporting tasks within a larger orchestration
  • Latency-sensitive, cost-sensitive production workloads that don't need a flagship reasoning model

How to access

ProviderModel ID
OpenAI ↗gpt-5.4-nano
OpenRouter ↗openai/gpt-5.4-nano

GPT Nano — every version

The full lineage of the GPT Nano line, newest first. Every version has its own page — click any to compare specs, benchmarks and pricing.

VersionReleasedContextLicense
GPT-5.4 nanocurrent2026-03-17Proprietary
GPT-5 nano2025-08-07Proprietary

FAQ

What is GPT-5.4 nano?

GPT-5.4 nano is the smallest and cheapest model in OpenAI's GPT-5.4 line, released on March 17, 2026. OpenAI describes it as "our cheapest GPT-5.4-class model for simple high-volume tasks" such as classification, data extraction, ranking, and coding sub-agents, where speed and cost matter more than deep reasoning.

How much does GPT-5.4 nano cost?

GPT-5.4 nano costs $0.20 per million input tokens and $1.25 per million output tokens, with cached input at $0.02 per million tokens on the standard tier. That makes it the lowest-priced model in the GPT-5.4 family — about four times the price of the older GPT-5 nano.

What is the context window and max output of GPT-5.4 nano?

GPT-5.4 nano has a 400,000-token context window and can generate up to 128,000 output tokens in a single response. Its knowledge cutoff is August 31, 2025.

How is GPT-5.4 nano accessed, and does it support images?

GPT-5.4 nano is available through the OpenAI API (model ID gpt-5.4-nano), as well as Codex and ChatGPT. It accepts text and image input and returns text, and supports reasoning, function calling, structured outputs, and streaming.