Google Adds Cheaper Gemini Models as Pro Stays Delayed

SAN FRANCISCO: Google has released three lower-cost Gemini models designed for faster, specialised and less demanding workloads. However, the company has not announced a new launch date for Gemini 3.5 Pro, its most anticipated premium model.

The new lineup includes Gemini 3.6 Flash, Gemini 3.5 Flash-Lite and Gemini 3.5 Flash Cyber. Google is positioning these models as practical options for organisations that do not need its most advanced technology for every task.

The update expands the number of affordable Gemini options available to businesses and developers. Still, attention remains fixed on Gemini 3.5 Pro, which was previously expected to launch in June and is now undergoing testing with selected partners.

Key Points

• Google has introduced three lower-cost Gemini models.

• Gemini 3.5 Flash Cyber is designed for cybersecurity applications.

• Gemini 3.5 Pro has missed its expected June launch window.

• Google has not provided a confirmed replacement release date.

• The flagship model is currently being tested with partners.

• Training has already started on Gemini 4.

Google Introduces Three Lightweight Gemini Models

The latest releases focus on users who require reliable performance without the higher cost and computing demands associated with a flagship model.

Gemini 3.6 Flash is an updated lightweight option developed for workloads where speed and cost matter more than access to the most advanced capabilities.

Gemini 3.5 Flash-Lite is another lightweight model intended for simpler and potentially high-volume tasks. It gives businesses an option for handling routine workloads without using a premium model for every request.

Google has also introduced Gemini 3.5 Flash Cyber, a specialised model designed for cybersecurity applications. Google DeepMind described it as a cost-efficient alternative to larger and more expensive cybersecurity models.

New Google Gemini Model Lineup

The lineup shows that Google is not relying on a single model for every customer. Instead, it is offering different options based on performance requirements, operational costs and specialised use cases.

Why Is Gemini 3.5 Pro Delayed?

Google CEO Sundar Pichai previously indicated that Gemini 3.5 Pro was planned for a June launch during his keynote address at the company’s I/O developer conference in May.

That launch window has now passed, and Google has not announced a revised date.

According to the report, Gemini 3.5 Pro has reportedly fallen short of certain internal performance targets, particularly in coding. This area has become increasingly important as companies adopt advanced models for software development, code review and other technical workloads.

Google says the model is being tested with partners and will arrive “soon.” However, the company has not explained how long the testing process may continue or when wider access will begin.

The delay is significant because Gemini 3.5 Pro is expected to represent Google’s strongest response to premium models offered by OpenAI and Anthropic.

Competition Places More Pressure on Google

Google strengthened its position in the technology market after releasing the Gemini 3 model family in November. The release was seen as an important step for Google DeepMind as competition among major model developers intensified.

Since then, Google’s rivals have continued updating their own products. Anthropic released its latest Mythos 5 and Fable 5 models, while OpenAI launched GPT-5.6 earlier in July.

This makes Gemini 3.5 Pro an important test for Google. Investors and enterprise customers will be watching whether the model can compete in areas such as:

• Coding performance

• Enterprise productivity

• Advanced problem-solving

• Operating costs

• Large-scale business deployment

A strong launch could help Google protect the momentum created by the Gemini 3 family. A longer delay, however, could give competitors more time to strengthen their positions in premium enterprise applications.

Google Highlights Gemini’s Cost Advantage

While users wait for Gemini 3.5 Pro, Google is placing greater emphasis on the affordability of its lighter models.

Pichai has said that some companies could save more than $1 billion annually by shifting most of their artificial intelligence workloads to Gemini models.

The estimate supports Google’s wider strategy of attracting organisations that need to control the cost of operating advanced technology at scale.

This approach is also visible in the positioning of Gemini 3.5 Flash Cyber. Rather than competing only through maximum performance, Google is presenting the model as a capable cybersecurity option that may cost less to operate than larger systems.

For many businesses, pricing could be as important as benchmark performance. A company processing large numbers of requests may prefer a faster and more affordable lightweight model for routine work while reserving premium models for complex tasks.

Gemini 4 Training Has Already Started

Google confirmed that training has begun on Gemini 4, even as Gemini 3.5 Pro remains in partner testing.

The disclosure shows that work on the company’s next model generation is already moving forward. However, the immediate focus remains on the delayed Pro release and whether it can meet Google’s internal expectations before reaching customers.

Alphabet’s second-quarter earnings conference call could bring further questions about the model’s performance and release schedule.

For now, Google has strengthened its lower-cost lineup with three new options. The more important development will be the arrival of Gemini 3.5 Pro, which could determine how effectively Google competes with OpenAI and Anthropic in the premium model market.

Source

This information is based on an ARY News report credited to Reuters and published on July 21, 2026.