Google: Gemini 3 Flash Preview API Price, Cost Calculator & Specs

At $0.50/$3 per 1M tokens, Gemini 3 Flash Preview is priced to be your workhorse model—not a premium planner, but a solid executor. The 6:1 output-to-input ratio means it's not the cheapest for long-form generation, but compared to Claude Haiku ($1/$5), it's about 40% cheaper across typical workloads. Its 1M context window and near-Pro reasoning make it a strong candidate for the "do the actual work" role in your architecture.

Best For:

Agentic Workflow Runner: Long-running agent loops that need to call tools, reason through steps, and keep context across many turns—this model won't bankrupt you while doing the heavy lifting.

Interactive Coding Assistant: Real-time code suggestions, debugging, and collaborative development where latency matters. The multimodal support (images, PDFs, video) lets you drop in a screenshot or document for context.

High-Volume API Traffic: If you're processing thousands of requests per day, the sub-dollar input price means your COGS stays manageable even at scale. Good for embedding, classification, and routine NLP tasks.