Overview
Gemini 3.1 Flash-Lite Preview is a low-latency, cost-efficient Gemini 3-series model designed for high-volume developer workloads. Google announced it on March 3, 2026, initially offering it in preview through the Gemini API in Google AI Studio and through Vertex AI. (blog.google)
Capabilities and uses
- Accepts text, images, video, audio and PDF inputs; produces text output.
- Supports configurable thinking, function calling, structured output, code execution, caching, file search, URL context, and grounding with Google Search and Google Maps.
- Suitable uses include high-volume translation, content moderation, image analysis and classification, data extraction, interface and dashboard generation, and agent workflows. (ai.google.dev)
Limits and availability
- Input limit: 1,048,576 tokens; output limit: 65,536 tokens.
- The preview model does not support image generation, audio generation, computer use or the Live API.
- Google says the preview model was shut down on May 25, 2026. Its recommended replacement is the separate model ID
gemini-3.1-flash-lite; the preview ID above is not currently available. (ai.google.dev)
Pricing
At launch, Google advertised USD $0.25 per 1 million input tokens and USD $1.50 per 1 million output tokens for Gemini 3.1 Flash-Lite. These are historical launch prices, not a currently available offer for the shut-down preview ID. The announcement does not specify modality-specific, cache or batch rates; those rates for this preview were not independently verified from an accessible official pricing page. The current official pricing page lists the separate stable model, so its rates should not automatically be treated as preview pricing. Pricing source checked November 16, 2026. (blog.google)