OpenAI has introduced its latest models, GPT 5.4 mini and nano, with GPT 5.4 mini now accessible for free to ChatGPT users. These new variants enhance coding, reasoning, multimodal understanding, and tool usage, performing over twice as fast as their predecessors. The mini model approaches the performance of the full GPT 5.4, excelling in evaluations like SWE-Bench Pro and OSWorld-Verified. Designed for latency-sensitive applications, these models empower responsive coding assistants and multimodal apps that interpret images in real-time. OpenAI highlights the importance of coding capabilities, as both GPT 5.4 mini and nano support workflows like targeted edits, codebase navigation, and debugging with low latency. Additionally, GPT 5.4 mini is now available in OpenAI’s API and Codex, while nano is exclusively accessible through the API, both at reduced costs, enhancing the versatility and affordability of AI-driven coding solutions.
Source link
