← All models

Gemini 2.5 Flash Preview

Published Publisher Googlegemini-2.5-flash-preview-09-2025

Overview

Gemini 2.5 Flash Preview is a multimodal preview model designed to improve quality and efficiency. Google announced it on September 25, 2025 for testing through Google AI Studio and Vertex AI; the release was not intended to become a new stable version. (developers.googleblog.com)

Capabilities and uses

  • Accepts text, images, video, and audio; generates text. The documented limits are 1,048,576 input tokens and 65,536 output tokens. Its knowledge cutoff is January 2025. (aistudio.google.com)
  • Supports thinking, function calling, code execution, structured output, Google Search grounding, URL context, context caching, file search, and the Batch API. These capabilities suit multimodal analysis, coding assistance, and multi-step agentic workflows. (aistudio.google.com)
  • Google reported improved agentic tool use and greater token efficiency with thinking enabled compared with the preceding release. (developers.googleblog.com)

Limitations and availability

The model does not support image or audio generation, the Live API, or Google Maps grounding. Google states that this exact preview model was shut down on February 17, 2026. (aistudio.google.com)

Pricing

The preview model's official historical input and output token prices could not be verified from the currently accessible Google pricing page. They are therefore unverified; pricing for the separately identified stable model should not be assumed to apply. Pricing source checked: Google Gemini API pricing, on November 18, 2026. (aistudio.google.com)

Official sources