OpenAI Unveils GPT-5.4 Mini and Nano: A New Era of Efficient AI

OpenAI has announced the release of GPT-5.4 mini and GPT-5.4 nano, the company’s latest advancements in large language models. These models are designed to deliver the power of GPT-5.4 in a faster, more efficient, and cost-effective package, targeting developer workflows where responsiveness and scalability are paramount. OpenAI officially introduced the models on March 18, 2026.

GPT-5.4 Mini: Powerful Reasoning for Production Workflows

GPT-5.4 mini is engineered to distill the strengths of GPT-5.4 into a smaller, more efficient model. It excels in coding, reasoning, multimodal understanding, and tool use, while achieving approximately 2x faster performance. Microsoft highlights its suitability for developer copilots and coding assistants, where low latency is crucial.

Key capabilities of GPT-5.4 mini include:

• Text and Image Inputs: Enables the creation of multimodal experiences combining prompts with visual data.
• Tool Use and Function Calling: Reliably invokes tools and APIs for agentic workflows.
• Web and File Search: Grounds responses in external or enterprise content for multi-step tasks.
• Computer Use: Supports software-interaction loops, interpreting UI state and taking actions.

GPT-5.4 Nano: Speed and Affordability for Simple Tasks

GPT-5.4 nano represents the smallest and most affordable variant, optimized for tasks like data sorting, information extraction, and supporting more complex workflows. According to OpenAI, this model embodies the industry trend towards specialized models working in concert to perform various tasks.

Availability and Access

GPT-5.4 mini is currently available in ChatGPT for both free and Plus users as part of the “Reflection” feature. GPT-5.4 nano is primarily accessible through APIs, catering to developers and companies building automation tools and systems. OpenAI’s API documentation provides detailed information for developers.

Technical Specifications

Here’s a breakdown of key technical details for GPT-5.4 mini:

• Context Window: 400,000 tokens
• Max Output Tokens: 128,000
• Knowledge Cutoff: August 31, 2025
• Pricing:
• Input: $0.75 per 1M tokens
• Cached Input: $0.075 per 1M tokens
• Output: $4.50 per 1M tokens

Regional processing endpoints incur a 10% uplift for GPT-5.4 mini.

API Endpoints Supported

GPT-5.4 mini supports a wide range of API endpoints, including:

• Chat Completions (v1/chat/completions)
• Responses (v1/responses)
• Realtime (v1/realtime)
• Assistants (v1/assistants)
• Batch (v1/batch)
• and more (see OpenAI’s API documentation for a complete list).

A Shift Towards AI Infrastructure

The release of GPT-5.4 mini and nano signifies a broader trend in the AI sector: moving beyond universal models towards a sophisticated infrastructure of specialized models. While large models demonstrate AI’s potential, smaller, faster, and more affordable models are crucial for widespread adoption and integration into everyday applications.