Gemini 2.5 Flash API - Pricing, Quickstart & Provider Comparison

What Is Gemini 2.5 Flash? Gemini 2.5 Flash is Google’s primary model for high-volume, latency-sensitive tasks that require reasoning. It’s the first Flash-class model with built-in thinking, a hybrid reasoning mode you can toggle on or off at will. That distinction makes it meaningfully different from 2.0 Flash and worth evaluating against models that cost significantly more. Gemini 2.5 Flash supports the following input types: text, code, images, audio, video, and documents. For document inputs, two constraints apply in production: maximum file size is 50MB per document (files exceeding this must be split into sub-50MB chunks before submission). Supported document MIME types are limited to application/pdf and text/plain only. What it does not support: audio generation, image generation, and the Live API. If you need image generation, use Gemini 2.5 Flash Image, which is a separate model. The thinking budget is a parameter that controls how much internal reasoning the model performs before generating a response. This is built into the model’s architecture during inference. Setting the budget to 0 disables it entirely, producing the fastest and cheapest output.…