Model positioning
Gemini 2.0 Flash-Lite is a multimodal model in the Gemini 2.0 family optimized for speed, scale, and cost efficiency. Google announced its general availability for production use on February 25, 2025. The earlier public preview was released on February 5, 2025. (aistudio.google.com)
Capabilities and uses
- Supports text, image, audio, and video inputs and produces text output; the input limit is 1,048,576 tokens, and the output limit is 8,192 tokens. (ai.google.dev)
- Supports function calling, structured output, context caching, and the Batch API. It can be used for multimodal content understanding, summarization, information extraction, and applications requiring long contexts or large-scale processing. (ai.google.dev)
- Google states that, compared with Gemini 1.5 Flash, the model has improved on reasoning, multimodal, math, and factuality benchmarks. (developers.googleblog.com)
Limitations and status
- Does not support image or audio generation, code execution, Google Search grounding, the Live API, or thinking mode. Outputs and factual claims may still contain errors; important results should be verified by a human. (ai.google.dev)
- This version was shut down on June 1, 2026, and is no longer available; Google recommends migrating to
gemini-3.1-flash-lite. (aistudio.google.com)
Pricing
The official Google Cloud Agent Platform price list still lists prices in US dollars (USD) for Gemini 2.0 Flash Lite, all billed per 1 million tokens: standard pricing is USD 0.075 for input tokens, USD 0.075 for input audio tokens, and USD 0.30 for output text tokens; Batch API prices are USD 0.0375, USD 0.0375, and USD 0.15, respectively. These are prices listed by that platform; they do not mean that the shut-down model can still be called, nor should they be taken directly as prices on other platforms. Date checked: November 19, 2026. (cloud.google.com)