Recent statements from Google DeepMind chief Koray Kavukcuoglu indicate the next flagship Gemini Pro model—Gemini 4—has entered early post-training after pre-training completion, with leadership aiming for release “as soon as possible” and well before year-end amid competitive pressure from OpenAI and Anthropic. This follows repeated delays to the unlaunched Gemini 3.5 Pro, as the team prioritized iterative Flash updates and shifted focus to the larger frontier effort. Post-training refinement, benchmark validation, and internal testing typically span weeks, creating uncertainty around an October window despite the September 26 timing. Trader sentiment reflects this ambiguity through dispersed probabilities across late-October dates and elevated odds on no release by month-end, underscoring how product timelines can slip even with strong internal momentum.