Google has reportedly delayed the launch of its advanced Gemini 3.5 Pro AI model from June to July 2026, according to sources familiar with the development. The company is said to be using the additional time to gather feedback from early testers and refine the model’s performance in coding, token efficiency, and long, multi-step tasks.
The postponement has not been officially confirmed by Google. A company spokesperson declined to comment when asked about the revised timeline. As a result, the July release window remains based on information from an unnamed insider.
Extended Testing Phase
Early users have been testing Gemini 3.5 Pro through Google’s Antigravity development environment and the LMArena benchmarking service. The company is reportedly collecting feedback and making adjustments before a wider release, particularly focusing on complex, multi-step workflows that require maintaining context over several stages.
Google introduced the Gemini 3.5 model family at its I/O developer conference in May. CEO Sundar Pichai stated that Gemini 3.5 Pro would arrive the following month, though no firm date was provided. The company first launched Gemini 3.5 Flash, a model optimized for coding, AI agents, and complex workflows. Gemini 3.5 Pro is expected to offer greater computing power for demanding tasks.
Feedback from Gemini 3.5 Flash Influences Pro Development
Reports indicate that Google has incorporated feedback from Gemini 3.5 Flash users into the Pro model’s development. Some users noted that Flash consumed tokens too quickly, potentially raising costs for long prompts or extended tasks. Google is reportedly examining this issue while preparing Gemini 3.5 Pro.
Token usage directly affects how much text a model processes and generates. Higher consumption can increase costs for developers and enterprises running large numbers of AI requests. The upcoming Pro model is also expected to improve performance on long-horizon tasks that require planning, tool use, and sustained context, such as software development, data analysis, and automated business processes.
However, Google has not published full technical details or benchmark results for Gemini 3.5 Pro. Claims about speed, efficiency, and coding abilities will remain unverified until official data is released or wider access is granted.
Competition in AI Coding Intensifies
The reported delay comes as major AI developers compete for business customers, especially in coding tools. OpenAI and Anthropic have also released models aimed at coding and agent-based tasks. Google already offers Gemini through AI Studio, Vertex AI, Android Studio, and Antigravity. Gemini 3.5 Pro could expand those services once available.
Earlier in June, Google introduced Gemma 4 12B, a lightweight multimodal model designed to run locally on laptops with 16GB of RAM or video memory. Unlike Gemini’s closed commercial models, Gemma provides open model weights for developers seeking more control over local deployment.
For now, Gemini 3.5 Pro has no confirmed public launch date. Reports point to July, but Google has not specified which products, users, or regions will receive access first.