Internal model IDs spotted in Google's development tools hint at interim releases while the flagship Gemini 3.5 Pro faces continued delays
Google appears to be prepping two new AI models for deployment, with identifiers for Gemini 3.6 Flash and Gemini 3.5 Flash Lite surfacing inside the company’s development infrastructure. The model ID gemini-3.6-flash-tiered was spotted in Google’s Antigravity IDE, suggesting the models are at least close to being ready for external access through AI Studio and the Vertex API.
What we actually know about the new models #
The Gemini 3.6 Flash and 3.5 Flash Lite registrations were first noted on July 21, 2026. Internal registration in development tools is typically one of the final steps before a model becomes publicly accessible, though it doesn’t guarantee an imminent launch.
Google’s existing Gemini 3.5 Flash model remains operational on both AI Studio and Vertex AI. That model currently runs at approximately $1.50 per million input tokens and $9 per million output tokens, with a context window of 1 million tokens.
The “tiered” suffix in the registered ID for the 3.6 Flash model hints at a pricing structure with multiple access levels, possibly differentiating between rate-limited free tiers and paid production tiers.
Neither model has received a formal public announcement from Google, and no documentation has appeared on the company’s developer pages as of this writing.
The Gemini 3.5 Pro problem #
Google’s Gemini 3.5 Pro has been conspicuously delayed. The Pro model represents Google’s top-tier reasoning and capability offering, designed to compete with the most advanced models from OpenAI, Anthropic, and other frontier labs. Releasing Flash and Flash Lite variants as interim solutions while the Pro model faces delays is consistent with Google’s historical pattern of pairing rapid Flash-tier releases with more refined Pro models.
What this means for the AI market and investors #
Google parent Alphabet remains one of the largest companies investing in AI infrastructure and model development globally. The registration of new model variants signals continued commitment to expanding the Gemini ecosystem, even as the flagship release timeline slips.
At roughly $1.50 to $9 per million tokens for the current Flash model, Google is positioning itself competitively against rivals. A Flash Lite model could push those costs even lower, potentially undercutting competitors who charge premium rates for comparable inference quality.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our