Gemini 3.1 Flash Lite (Vertex AI, US)

Gemini 3.1 Flash Lite sits at the bottom of the Gemini 3 lineup as a Flash-Lite variant aimed at narrow, high-volume jobs that orbit a larger model rather than general chat. According to available comparisons, it is tuned specifically for translation, transcription, structured extraction, and request routing — the kinds of tasks where low latency and predictable cost matter more than frontier reasoning. Both comparison sources frame it as Google's most cost-efficient Gemini 3 release, positioned at roughly half the input and output price of Gemini 3 Flash, which makes it attractive for production pipelines that process millions of tokens a month. It also keeps the family's roughly one-million-token context window while sitting at the cheapest tier, so it can hold long documents for extraction or summarization without the per-token penalty of larger siblings.

Beyond raw cost, the model is also shaped for deployment flexibility. Routing data shows it exposed across multiple providers and regions on Vertex, with the EU region reported as the greenest footprint at around 2.41 gCO₂e per 1k tokens versus 4.24 in its highest-carbon region, giving cost-sensitive teams an additional lever beyond price. Practically, this makes Flash Lite a strong fit for builders who want to wire up a thin, fast layer underneath a heavier model — translating inputs, pulling structured fields out of long documents, transcribing audio, or deciding which downstream model should handle a request. It is less suited as a stand-alone answer engine for open-ended reasoning, but well matched to the connective, repetitive workloads where consistency, speed, and per-call price dominate the decision.