Key facts
- Google has delayed the release of its frontier AI model, Gemini 3.5 Pro.
- Rivals like Meta and OpenAI have released new top-tier AI models recently.
- Google's recent launches focus on faster, more cost-effective AI models.
- Analysts suggest Google's delay may shift its perception from leading-edge to trailing-edge in AI.
- Google's core businesses like Search and YouTube are experiencing strong momentum due to AI advancements.
Google is facing criticism and commentary from rivals regarding the delayed release of its frontier AI model, Gemini 3.5 Pro, even as the company rolls out faster and more cost-effective alternatives. Competitors like Meta and OpenAI have recently launched their own advanced models, leading some to question Google's position in the rapidly evolving AI landscape.
Alexandr Wang, Meta's chief AI officer, posted "gemini who?" on X in response to a leaderboard that placed Meta's Spark model above a recent Google launch. Similarly, Thibault Sottiaux from OpenAI replied to an announcement about Gemini 4 pre-training with "Hope it finishes one day too!" Google has declined to comment on these remarks.
Analysts suggest that the delay of Gemini 3.5 Pro has shifted the perception of Google's AI development from the "leading edge" to the "trailing edge." However, Google's core businesses, including Search, YouTube, and Cloud, continue to show strong momentum, partly due to AI advancements. The company is also emphasizing efficiency and cost-effectiveness in its model development, a strategy that has garnered praise from some users for specific applications like document extraction.
Despite these efforts, the trade-off between model efficiency and raw power is expected to be a topic during Google's upcoming Q2 earnings call. Analysts like Anshel Sag note that while Google's larger structure may lead to different development paces compared to competitors, it remains "too early to count anyone out" in the AI race.