Google Releases Nano Banana 2 Lite to GA and Gemini Omni Flash to Public Preview

Google on Monday made gemini-3.1-flash-lite-image (Nano Banana Lite / Nano Banana 2 Lite) generally available and launched Gemini Omni Flash in public preview, expanding its lineup of fast, cost-efficient generative media tools.

The updates arrive alongside the shutdown of older Veo video models and give developers immediate access to higher-volume image generation and a new natively multimodal video editing model.

Release Data

• Nano Banana 2 Lite (gemini-3.1-flash-lite-image): Now GA via Gemini API and Google AI Studio. Optimized for ultra-low latency image generation and editing in high-volume interactive and production workloads.

• Generates text-to-image outputs in under 4 seconds at $0.034 per 1K-resolution image. Positioned as the efficiency-focused model in the Nano Banana family for bulk content creation, consistent frame generation, and high-frequency pipelines.

• Gemini Omni Flash: Public preview for natively multimodal video generation and conversational editing. Supports up to 10-second 720p videos with multimodal inputs. Priced at $0.10 per second. Uses the Interactions API to maintain session context for up to three sequential natural-language edits (scene changes, style transfers, product swaps). Can be chained with image models such as Nano Banana 2 Lite.

• Availability: Both models live today in Google AI Studio, Gemini API, and Gemini Enterprise Agent Platform. Nano Banana 2 Lite is rolling out more broadly to Gemini app, NotebookLM, Google Photos, Search/AI Mode, and other Google surfaces.

The releases reflect Google’s push to deliver specialized, high-throughput models alongside its frontier systems. Nano Banana 2 Lite targets cost-sensitive, high-volume image workloads that previously required heavier models, while Omni Flash introduces conversational, multi-turn video editing grounded in multimodal understanding — a step beyond single-shot generation.

For high-intent readers evaluating generative AI infrastructure and content platforms, these launches lower the cost and friction of scaling image and short-form video production.

The sub-4-second image generation at aggressive pricing and the ability to perform multiple contextual video edits in one session create new operational leverage for marketing teams, social platforms, and AI-native creative tools.

Combined with today’s Veo model sunsets, Google is actively migrating developers toward faster, cheaper, and more interactive media generation primitives across its ecosystem.