Google is set to launch its most affordable AI model yet, the Gemini 1.5 Flash 8B, for commercial use, according to a recent report from technology media outlet NeoWin. This announcement comes as the company aims to broaden accessibility to its AI capabilities.
The Gemini 1.5 Flash 8B model, which features 8 billion parameters, is a smaller version of the originally introduced Gemini 1.5 Flash model. It has been specifically designed for multi-modal tasks, such as handling extensive data and providing long text summaries. Compared to its predecessor, the Gemini 1.5 Flash 8B offers reduced latency, making it particularly suitable for applications like chatbots, transcription services, and lengthy text translations.
One of the standout features of the Gemini 1.5 Flash 8B model is its budget-friendly pricing structure. The new rates are set to take effect on October 14. Under a context window of fewer than 128K tokens, the cost for input prompts will be $0.0375 per million tokens, while the output prompts will cost $0.15 per million tokens. Additionally, caching tokens will be available for just $0.01 per million tokens.
For context, the original Gemini 1.5 Flash model priced its output prompts at $0.30 per million tokens starting August 12, 2024. This change signifies a dramatic reduction in costs for users, with the new Gemini 1.5 Flash 8B costing half as much as its predecessor for output tokens.