Gemini 3.5 Pro improvement work delays its launch

Google Gemini 3.5 Pro launch delayed due to the work that the technology company is doing to improve its programming capabilities, to bring it up to par with the most recent and advanced models of OpenAI and Meta.

The tech company announced Gemini 3.5 Pro in May at its annual developer event, along with the faster variant, Gemini 3.5 Flash, with the promise of its most capable model would arrive in June.

Now in mid-July it has been confirmed that it will arrive later than expected as a result of the problems that Google is facing to improve your capabilities, especially in programmingas sources familiar with this matter have indicated to Bloomberg.

As part of these improvement works, Google updated last month the database which it uses to train Gemini to improve its ability to write code, although without much success, in an attempt to catch up with its two main competitors in this field, OpenAI and Meta.

Precisely, both companies have launched new models, occupying the silence of Gemini 3.5 Pro. On the one hand, OpenIA now has available family GPT-5.6 (Sol, Luna, Terra) which, in addition to offering outstanding agentic programming capabilities, comes with important cybersecurity protections.

Meta, for his part, emphasizes the capabilities of Muse Spark 1.1 in personal agent tasks and to address complex projects much faster, in addition to your programming work, with improvements in diagnosing and correcting errors, deploying functions in enterprise systems, and performing large-scale code migrations.

“We are currently testing the 3.5 Pro, an improved version of the Flash model, and other models in collaboration with our partners; Likewise, we maintain a constructive collaboration with the US government regarding model testing and the establishment of broader frameworks for action,” a Google spokesperson explained to Bloomberg in a statement reported by Europa Press.