Google Rolls Out Gemini 2.5 Flash, a Cost-efficient AI Model | Beebom
Google has recently introduced Gemini 2.5 Flash, an AI model that aims to provide a cost-efficient solution for developers. This new model is currently in preview and can be accessed for free on the Gemini app. Developers are now able to utilize the API on AI Studio and Vertex AI.
Gemini 2.5 Flash: A Hybrid Reasoning Model
According to Google, Gemini 2.5 Flash is the company's "first fully hybrid reasoning model." This unique feature combines the characteristics of a traditional LLM with a reasoning/thinking AI model. Developers have the flexibility to toggle thinking on or off and allocate thinking budgets as needed.
Moreover, the AI model is optimized for cost, quality, and latency, striking a balance between fast performance and high-quality output at a lower cost. While Gemini 2.5 Pro offers top-notch performance, Gemini 2.5 Flash prioritizes affordability.
Benchmarks and Performance
In terms of benchmarks, Gemini 2.5 Flash slightly lags behind OpenAI's latest o4-mini reasoning model. On Humanity's Last Exam, Gemini 2.5 Flash achieves a score of 12.1%, whereas o4-mini scores 14.3%. Similarly, on GPQA Diamond, Gemini 2.5 Flash reaches 78.3%, while o4-mini performs slightly better at 81.4%. Additionally, on AIME 2025, Gemini 2.5 Flash scores 78% compared to o4-mini's 92.7%.
Affordable Pricing
The true value of Gemini 2.5 Flash lies in its competitive pricing. For non-reasoning API calls, the input/output cost is $0.15/$0.60, while the cost for reasoning API calls is $3.50. Additionally, Gemini 2.5 Flash is compatible with new features like Canvas in the Gemini app, enhancing its usability and functionality.
Passionate about Windows, ChromeOS, Android, security, and privacy issues. Have a penchant to solve everyday computing problems.
