Google Upgrades Gemini With New Tools for Developers

Google is on a spree to launch new features for its Gemini model. Recently, the company introduced support for implicit caching in Gemini 2.5 models and the addition of image generation and editing tools in Gemini 2.0 Flash. The new features primarily focus on saving costs and giving developers creative capabilities. So, let’s take a closer look at everything new at Gemini. 

Implicit Caching Now Supported in Gemini 2.5 

Google introduced explicit caching in 2024 and has now rolled out implicit caching support across Gemini 2.5 models. This update allows developers to benefit from token cost savings of up to 75% without having to manually manage cache creation.

With implicit caching, the Gemini API detects when incoming prompts share a common prefix with previous requests and automatically applies a discounted rate for those repeated tokens. Developers are encouraged to keep stable, prompt content at the beginning of their requests to increase the likelihood of cache hits.

Apart from that the tech giant has also lowered the minimum request size needed to qualify for caching: 1024 tokens for Gemini 2.5 Flash and 2048 tokens for Gemini 2.5 Pro. Explicit caching options remain available for those seeking guaranteed cost savings. 

Image Generation Preview Available in Gemini 2.0 Flash 

In a separate update, Google has launched image generation and editing capabilities for Gemini 2.0 Flash in preview. This feature, now accessible via Google AI Studio and Vertex AI, allows developers to generate and modify images using natural language prompts.

The image generation tool offers improvements over earlier experimental versions, including better visual quality, improved accuracy in text rendering within images, and fewer blocked outputs due to content filters. The preview also comes with higher rate limits and updated pricing.

Developers can start using the feature by accessing the “gemini-2.0-flash-preview-image-generation” model name in supported platforms. 

On a similar note, the company has also launched its Gemini 2.5 Pro (I/O edition), which is in early access. It is claimed to improve Gemini’s performance across coding tasks, code editing, and even complex agent workflows. 

Looking Ahead  

That’s it for now. As Google I/O 2025 approaches, we can expect more beta releases and early access updates for upcoming features. Like previous years, the biggest announcements will likely be saved for the main event. However, what’s clear, though, is that Gemini is set to receive massive improvements in the coming days.