OpenAI Developers

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 version of GPT-5.4. openai.com/index/introduc…

You can now summon Codex from anywhere on your desktop. It's easier than ever to reach for Codex.

@yashrajnayak ❤️

Thank you everyone who joined us last Saturday for India’s first official #Codex Meetup in Bengaluru 🇮🇳 Many more to come! You can just build things ✨

Find a Codex meetup next to you: developers.openai.com/community/meet…

Codex Hackathon Delhi 📆 April 18 📍 Delhi 🇮🇳 luma.com/rjsbm9qw

Codex is coming to India 🇮🇳 Join events led by Codex ambassadors to: Get hands-on with Codex Meet developers building with Codex Ship projects at hackathons

@zapier @OpenAI 🤝

GPT-5.4 Mini & Nano are live in Zapier. We ran the benchmarks 🧵 Both of @OpenAI's new models matched Haiku 4.5's performance at roughly half the cost. Mini on low reasoning, Nano on medium reasoning. OpenAI built these models specifically for high-volume, latency-sensitive work: coding assistants, subagents, image processing. Fast by design, not by compromise. Building at scale? Need speed without blowing your budget? Try them in Zapier now.

Are you up for a challenge? openai.com/parameter-golf

ICYMI x.com/OpenAIDevs/sta…

We’re reviewing Codex for Open Source applications on a rolling basis. Here are some of the projects and maintainers who just joined our Codex for OSS community 👇 x.com/athasdev/statu…

@gabrielchua ❤️

Now with `gpt-5.4-mini` and `nano` out, I put together a simple cheat sheet of the latest OpenAI models by use case. Noticed at a few recent hackathons & meetups: some folks still default to `gpt-4o-mini` for LLMs and `whisper-1` for transcription. Newer options tend to fit better now with much better performance. If you’re running into issues switching, lmk!

GPT-5.4 mini matters for subagents because it changes what feels worth handing off. The parent thread should hold the architecture, plan, and progress narrative. Fast subagents can explore the repo, check hypotheses, and preserve the parent thread’s limited attention.