Google Launches Gemini 2.5 Pro and Flash AI Models, Adds Low-Cost Flash-Lite Option

Google has announced that its Gemini 2.5 Pro and Gemini 2.5 Flash AI models are now ready for enterprise production after a period of experimental preview. The company also brought a new variant, Gemini 2.5 Flash-Lite, which is called the most cost-effective option in its lineup for high-volume tasks.

The Gemini 2.5 models are said to have “reasoning” capabilities, which will let developers control how much computational effort the model spends on complex problems. This “thinking budget” means businesses can tune the AI for either accuracy or speed, depending on the task, as reported by VentureBeat.

Gemini 2.5 Pro is said to be Google’s most capable model. It can process up to one million tokens of context, which is about 750,000 words. This lets it analyze entire codebases or lengthy documents in a single session. Gemini 2.5 Flash is for high-throughput tasks ...