Google Gemini 2.5 Flash

The workhorse model: Gemini 2.5 Flash delivers exceptional speed and cost-efficiency with robust multimodal and reasoning capabilities.

Gemini 2.5 Flash is Google's most efficient, high-speed workhorse model, engineered for low-latency, high-volume tasks. It significantly improves upon its predecessor, using 20-30% fewer tokens in evaluations while boosting performance across key benchmarks (reasoning, multimodality, code). This model is fully multimodal, accepting inputs including Text, Code, Images, Audio, and Video. A key feature is its new 'thinking' capability, which allows developers to control the model's reasoning budget to precisely balance output quality, latency, and cost for specific applications. It is available via the Gemini API in Google AI Studio and Vertex AI.

https://cloud.google.com/vertex-ai/docs/generative-ai/model-reference/gemini-2.5-flash