OpenAI Launches Ultra-Fast GPT-5.4 Mini and GPT-5.4 Nano Models

Two New Speed-Optimized Models

OpenAI has released two new AI models — GPT-5.4 mini and GPT-5.4 nano — both engineered for scenarios where response latency is a critical factor.

«We're introducing GPT-5.4 mini and nano, our most capable small models yet. GPT-5.4 mini is more than 2x faster than GPT-5 mini. Optimized for coding, computer use, multimodal understanding, and subagents. For lighter-weight tasks, GPT-5.4 nano is our smallest and cheapest…» — OpenAI Developers (@OpenAIDevs), original post

According to the company, GPT-5.4 mini substantially outperforms GPT-5 mini in code generation, logical reasoning, multimodal analysis, and tool integration. Processing speed has increased by more than 2x. On benchmarks such as SWE-Bench Pro and OSWorld-Verified, the model approaches the performance levels of the full GPT-5.4.

Why This Matters

The launch of GPT-5.4 mini and nano reflects an industry-wide push toward compact yet powerful models that operate in real time. For developers building AI-powered applications — from chatbots to automated agents — response latency directly determines user experience and product viability. Making these models available via API enables developers to build fast, responsive services without relying on expensive full-scale neural networks.

OpenAI highlighted key use cases where the new models excel:

• AI assistants with instant response times;
• Subagents that quickly handle auxiliary tasks;
• Interface management systems that capture and interpret screenshots;
• Multimodal applications for real-time image processing.

The company emphasized that for these scenarios, the best model is not necessarily the most powerful one — it's the one that responds instantly, reliably interacts with tools, and still handles complex professional tasks effectively.

GPT-5.4 nano is the smallest and fastest variant in the GPT-5.4 lineup. OpenAI recommends it for data classification and extraction, ranking, and subagents tasked with simple auxiliary operations.

Coding and Development

OpenAI specifically highlighted GPT-5.4 mini's effectiveness in software development workflows where rapid iteration speed is essential. The model handles targeted code edits, codebase navigation, frontend generation, and debugging cycles with minimal latency.

In benchmarks, GPT-5.4 mini outperforms GPT-5 mini at comparable response speeds, while operating significantly faster in several tests.

Computer Use

GPT-5.4 mini delivers strong results in multimodal tasks related to interface management. The model can instantly interpret screenshots of complex user dashboards and execute computer actions based on that analysis.

Availability and Pricing

GPT-5.4 mini is available through the API, Codex, and ChatGPT. Pricing is set at $0.75 per million input tokens and $4.5 per million output tokens.

GPT-5.4 nano is currently available exclusively through the API at lower rates: $0.2 per million input tokens and $1.25 per million output tokens.