Gemini 3.5 Pro delay: what Google’s behind-schedule AI coding tool race means for builders

Google's next flagship AI model, Gemini 3.5 Pro, is reportedly months behind schedule because its coding capabilities have not met internal targets. The delay, first reported by Bloomberg citing 10 current and former employees, reflects deeper structural problems inside Google that have slowed its ability to compete with OpenAI and Anthropic. For AI builders and engineering teams, the news matters because it affects planning for AI-assisted coding workflows and raises questions about how to evaluate Google’s model pipeline against competitors that are shipping faster.

What happened

Google is months behind on delivering Gemini 3.5 Pro, the planned upgrade to its flagship model. The company was widely expected to release the model at its May developer conference but has not been able to close the gap with Anthropic and OpenAI in writing code, according to the report. A late attempt to improve coding by updating training data also produced disappointing results.

Part of the problem is internal fragmentation. Google Cloud, DeepMind, and the Android team are all building AI coding tools for developers, creating internal competition that has slowed progress. Co-founder Sergey Brin has pushed for faster AI coding, but his efforts have been hampered by competing factions and by engineers who believe important code should still be written by humans.

Google has taken steps to consolidate. Chief AI Architect Koray Kavukcuoglu is working to unite internal AI coding tools, and a new DeepMind team led by research engineer Sebastian Borgeaud has been formed to tackle the problem. The company said at its most recent Cloud conference that 75 percent of code at Google is now AI-generated and that most developer tooling has been consolidated under Antigravity, an internal platform that manages data, memory, and safety protocols for AI applications.

The delays have contributed to senior departures to Anthropic and other labs, with former employees citing frustration with Google’s competitive position. Engineers also face capacity constraints on computing power, even for internal use. Only some teams inside Google are allowed to use Anthropic’s Claude, with access restricted to groups doing cutting-edge research.

Customer experiences with the current Flash model have been mixed. Rodrigo Davies, a product manager at Figma, said the model hit “a sweet spot of speed and quality” for the design platform’s AI assistant. But Freddy Vega, CEO of Platzi, said the Flash model is more expensive and slower than its predecessor while remaining far less capable than competitors, and his team has shifted to Anthropic.

Why AI builders should care

For teams building AI-assisted coding workflows or planning to integrate Gemini Pro into their stack, the delay means uncertain timelines. If Google’s next flagship model is months late, enterprise adoption plans that depended on it may need to hedge. The internal fragmentation is a warning: when multiple teams inside a vendor build overlapping AI tools, it can stall progress and lead to inconsistent quality.

Competitors are not standing still. OpenAI, Meta, and Anthropic have all released models that outperform Google’s current coding capabilities, making it harder for Google to catch up. For AI builders, this means the choice of model provider may shift based on who can ship reliable coding tools first.

Practical implications

Engineering teams should treat Gemini 3.5 Pro as an uncertain dependency. If your product roadmap relies on its coding performance, build in fallback options using alternative models, including Claude, GPT-4o, or open-source models. Watch for pricing and performance changes in the Flash model line, which is Google’s current shipping alternative for low-latency use cases.

Google’s consolidation efforts under Antigravity and the new DeepMind team are worth monitoring. If they succeed, Google could ship a more unified and capable coding toolset. But the current signal is that internal coordination remains a bottleneck.

Caveats

All reports about the delay and internal dynamics come from unnamed sources and industry reports. Google’s official statement says it is “shipping quickly across a wide range of models” and testing the upgraded Pro, a new Flash model, and other models with partners. Customer experiences with the Flash model vary significantly by use case, so your own experience may differ.