{"id":2165,"date":"2026-07-22T10:12:34","date_gmt":"2026-07-22T10:12:34","guid":{"rendered":"https:\/\/techvisor.pro\/?p=2165"},"modified":"2026-07-22T10:12:34","modified_gmt":"2026-07-22T10:12:34","slug":"gemini-ai-the-new-3-6-flash-models-are-out-but-the-pro-version-has-been-delayed-again","status":"publish","type":"post","link":"https:\/\/techvisor.pro\/en\/gemini-ai-the-new-3-6-flash-models-are-out-but-the-pro-version-has-been-delayed-again\/","title":{"rendered":"Gemini AI: The new 3.6 Flash models are out, but the Pro version has been delayed again"},"content":{"rendered":"

On July 21, 2026, Google DeepMind released three new Gemini AI models at once \u2014 but not the one everyone had been waiting for. We break down exactly what was released, why it matters for developers, and why the flagship Gemini 3.5 Pro still has not appeared.<\/p>\n

What Exactly Google Released: Three New Models<\/h2>\n

All three models belong to the \u201clighter\u201d Flash class \u2014 Google is deliberately developing this tier for fast and inexpensive AI agents, rather than for maximum depth of reasoning.<\/p>\n

\"Gemini<\/h3>\n

 <\/p>\n

 <\/p>\n

This is the main model in the lineup, focused on code, knowledge work and multimodal tasks. The main feature is efficiency: the model uses 17% fewer output tokens compared with the previous 3.5 Flash, and on some coding benchmarks the savings reach 65%. The price has also gone down \u2014 $1.50 per million input tokens and $7.50 per million output tokens, compared with $9 for output tokens in the previous version.<\/p>\n

Gemini 3.5 Flash-Lite \u2014 the Fastest and Cheapest<\/h3>\n

This model is aimed at high-volume tasks under heavy load: 350 output tokens per second, with pricing starting at $0.30 per million input tokens. It is an ideal option for tasks where speed and cost matter more than depth of analysis.<\/p>\n

Gemini 3.5 Flash Cyber \u2014 a Cybersecurity Model With Limited Access<\/h3>\n

\"Gemini<\/p>\n

A specialized model trained to find and fix vulnerabilities in code works inside Google\u2019s CodeMender agent. Unlike the two previous models, access to Flash Cyber is limited \u2014 only governments and verified partners within a pilot program, due to the dual-use risks of this technology.<\/p>\n

Why Gemini 3.5 Pro Still Has Not Been Released<\/h2>\n

The most interesting part of this release is not what came out, but what did not. The flagship Gemini 3.5 Pro model, which has been expected for several months, once again did not appear in public access. According to Google\u2019s official wording, the model is in testing with partners, and the company plans to make it broadly available \u201cas soon as it is ready\u201d \u2014 without a specific date.<\/p>\n

Bloomberg reported internal performance-related delays with the model. Google did not directly deny this information, limiting itself to a general comment about partner testing.<\/p>\n

What Comes Next: Google Is Already Starting to Train Gemini 4<\/h2>\n

\"Google<\/p>\n

In parallel with explanations about the Pro delay, Google reported something unexpected: the company has already begun preliminary training of Gemini 4, calling the process the \u201cmost ambitious training cycle\u201d in the history of the lineup. This means that while Gemini 3.5 Pro is stuck at the testing stage, the company\u2019s resources are already partly directed toward the next generation of models.<\/p>\n

In Brief: The Key Points<\/h2>\n