Several developers recently discovered an unreleased new large model "GPT-5.6" in the backend logs of OpenAI Codex. This new flagship, internally named "iris-alpha," has shown astonishing technical specifications and is expected to be officially released in June this year.

1.5 Million Token Context Window Breaks Records
In the GPT-5.6 series model that was exposed, the most impressive breakthrough is its ultra-long context window of up to 1.5 million tokens. Compared with the current GPT-5.5 API, the new model's maximum text processing limit has increased by nearly 43%, which means it can easily handle long contracts, large codebases, and complex multi-step long-term tasks.
Developers tested it in the auxiliary tool OpenCode. The results confirmed that when the input data reached 900,000 tokens, the model could still respond smoothly, and even perfectly handle overload requests exceeding 1.05 million tokens. Also exposed are two mysterious versions called ember-alpha and beacon-alpha, but their specific positioning remains unclear.

Commercial-Level Frontend Generation Causes Industry Shock
Aside from its powerful text capacity, GPT-5.6 also brings a revolutionary evolution in frontend interface generation. Leaked screenshots show that without any detailed prompt, the model directly generated a minimalist note-taking application interface called Lumen Notes.
