← All models

Gemini 2.5 Flash-Lite Preview

Published Publisher Googlegemini-2.5-flash-lite-preview-06-17

Model positioning

A preview version of Gemini 2.5 Flash-Lite, intended for tasks sensitive to cost, latency, and throughput. At launch, Google positioned it as the lowest-cost, lowest-latency model in the Gemini 2.5 series. (developers.googleblog.com)

Capabilities and uses

  • Supports multimodal input and has a context window of approximately 1 million tokens. (blog.google)
  • Supports adjusting the thinking budget through an API parameter; thinking is disabled by default to optimize speed and cost. (developers.googleblog.com)
  • Supports function calling and tools such as grounding with Google Search, code execution, and URL Context. (developers.googleblog.com)
  • Suitable for high-throughput or low-latency scenarios such as large-scale classification, summarization, and translation. (developers.googleblog.com)

Limitations and status

This is a preview version; Google positions Flash-Lite for scenarios with relatively lower model capability requirements that are sensitive to cost and latency. On November 4, 2025, Google announced that this preview model was scheduled to shut down on November 18, 2025; the preview API should not be assumed to remain available. (developers.googleblog.com)

Release date

June 17, 2025. Google's release notes explicitly list the API ID for this preview version. (aistudio.google.com)

Pricing

The official historical input and output token charges in US dollars (USD) for this preview version are unverified; prices for other versions cannot be treated as the price of this preview version. Official pricing lookup date: November 18, 2026. (aistudio.google.com)

Official sources