Google Gemini 3.1 Flash-Lite Targets Enterprise Scale at $0.25 Per Million Tokens

Google has launched Gemini 3.1 Flash-Lite in preview, the fastest and most cost-efficient model in the Gemini 3 family. Priced at $0.25 per million input tokens and $1.50 per million output tokens, Flash-Lite is designed for the high-volume, cost-sensitive workloads that make up the vast majority of enterprise LLM traffic — translation, content moderation, data extraction, and UI generation.

Speed and Cost

Flash-Lite delivers 2.5x faster time-to-first-token compared to its predecessor Gemini 2.5 Flash, with a 45% increase in output generation speed. For enterprises processing millions of requests per day, these improvements translate directly into lower infrastructure costs and better user experiences. At $0.25 per million input tokens, Flash-Lite is among the cheapest frontier-adjacent models available from any major provider.

The model is based on the Gemini 3 Pro architecture but with aggressive optimizations for latency and throughput. Google has not disclosed the exact parameter count, but the model's Arena.ai Leaderboard Elo score of 1432 places it comfortably above GPT-4 class models on conversational tasks — impressive for a model optimized primarily for speed and cost rather than peak capability.

Enterprise Use Cases

Google is positioning Flash-Lite as the default model for enterprise applications where cost and latency matter more than maximum reasoning ability. Specific use cases highlighted in the announcement include real-time content moderation for social platforms, high-volume document translation, UI component generation from design specifications, and simulation workloads that require thousands of parallel model calls.

The model achieves 86.9% on GPQA Diamond and 76.8% on MMMU Pro — benchmarks that measure scientific reasoning and multimodal understanding respectively. While these scores don't match the top-tier Gemini 3 Ultra model, they represent a dramatic improvement in the capability-per-dollar frontier.

Competitive Landscape

Flash-Lite enters a crowded market for efficient LLM inference. Anthropic's Claude 4.5 Haiku, OpenAI's GPT-5.4 Mini, and Meta's Llama 4 Scout all compete in the cost-efficient segment. Google's advantage is deep integration with its cloud platform — Flash-Lite is available through both Google AI Studio for experimentation and Vertex AI for production deployment, with seamless scaling and enterprise support.

Gemini 3.1 Flash-Lite is available immediately in preview through the Gemini API and Vertex AI.