Google has expanded its Gemini Enterprise Agent Platform with the launch of Nano Banana 2 Lite, a new AI image generation and editing model designed to help businesses create visuals more quickly while reducing operational costs. Alongside it, the company has also introduced Gemini Omni Flash in public preview, bringing conversational video generation and editing capabilities to developers.While both models are aimed at enterprise users, Google is positioning Nano Banana 2 Lite as the faster and more cost-efficient option within the Nano Banana family. The company says the model is built for scenarios where developers and businesses need to generate, edit and iterate on images at scale without compromising significantly on quality.Key TakeawaysGoogle has launched Nano Banana 2 Lite for the Gemini Enterprise Agent Platform.The new model focuses on faster and more cost-efficient AI image generation and editing.Google has also introduced Gemini Omni Flash in public preview for AI-powered video creation and editing.Nano Banana 2 Lite improves character consistency, text rendering and contextual image generation over its predecessor.Both models include SynthID watermarking and C2PA content credentials by default.Nano Banana 2 Lite targets speed and lower costs for AI image generationWith Nano Banana 2 Lite, Google is aiming to simplify image creation for businesses that rely on AI-powered visual content. The model is designed for rapid idea generation, A/B testing, marketing creatives, social media content, storyboarding and large-scale image editing workflows.According to Google, Nano Banana 2 Lite is the fastest and most affordable image generation model in the Nano Banana lineup. The company says it enables developers to move from an idea to a visual concept quickly, making it suitable for applications where multiple image variations need to be created in a short span of time.The model is now generally available through the Gemini Enterprise Agent Platform.Brings improvements over the previous Nano Banana modelGoogle says Nano Banana 2 Lite introduces several improvements over the earlier Nano Banana (Gemini 2.5 Flash Image) model.One of the biggest upgrades is its improved understanding of real-world context, allowing it to generate more accurate scenes, location-specific mock-ups and data visualisations. The company also says the model delivers better character consistency, helping maintain the appearance of people and objects across multiple image generations - an important feature for storyboards, virtual try-on experiences and creative projects.Another area of improvement is text rendering and localisation. Nano Banana 2 Lite can generate more legible text within images, making it easier for developers and designers to create advertising creatives, product visuals and multilingual marketing assets.Gemini Omni Flash brings conversational video editingAlongside Nano Banana 2 Lite, Google has unveiled Gemini Omni Flash in public preview. Unlike Nano Banana, which focuses on images, Gemini Omni Flash is designed for video generation and editing.The model allows users to edit videos using natural language prompts, enabling tasks such as swapping characters, changing lighting conditions, modifying camera angles and adding new objects while preserving the original audio and video tracks.Google says Gemini Omni Flash also supports multimodal inputs, allowing developers to combine text, images and videos to guide content generation. Every generated video includes native audio output, while the model is designed to maintain consistency in characters, objects and visual styles across edits.The company added that support for additional capabilities - including audio references, video references, scene extension, last-frame editing and higher-resolution outputs - will be introduced in a future update to the Gemini Enterprise Agent Platform API.Built around Gemini's world knowledgeGemini Omni Flash isn’t just another flashy video generator, at least by Google’s account. They say they built it around Gemini’s grasp of history, science, culture, and even the rules that shape the real world. So, videos from this model aren’t just lifelike - they actually make sense in context. You get scenes, characters, and objects that fit together, acting the way you’d expect in the real world. For developers, that means they can roll out videos with a layer of accuracy and consistency that feels pretty natural.Something else stands out: Gemini Omni Flash can put clear, readable text and graphics right into its videos. So whether you’re adding captions, stylish moving typography, or syncing on-screen text with what’s happening, everything comes together smoothly. This makes a big difference if you’re producing explainers, educational clips, product videos, or ad campaigns where the visuals and words have to do their job together.Google’s not stopping here. Down the line, they plan to let developers use audio and video references, tweak scenes with last-frame editing, stretch out the action, or bump up resolution - all through the Gemini Enterprise Agent Platform API. It’s about giving people more tools and flexibility to shape videos that really land.Enterprise-focused safety features included by defaultGoogle says both Nano Banana 2 Lite and Gemini Omni Flash include C2PA content credentials and SynthID invisible watermarks by default to help users identify AI-generated content and improve transparency.To support enterprise-scale deployments, Google has also introduced Provisioned Throughput (PT) for Nano Banana 2 Lite, enabling organisations to handle high-volume API requests more reliably. Similar support for Gemini Omni Flash is expected to roll out in the future.The company also highlighted that several partners, including Artlist, Figma and Manus AI, have already started testing Nano Banana 2 Lite for use cases ranging from creative design workflows to real-time image generation for presentations and websites.Nano Banana 2 Lite vs Gemini Omni FlashFeatureNano Banana 2 LiteGemini Omni FlashPrimary purposeAI image generation and editingAI video generation and editingAvailabilityGenerally availablePublic previewKey focusSpeed and cost efficiencyConversational video editingInput supportImages and textText, images and videoImprovementsCharacter consistency, better text rendering, world knowledgeAudio generation, multimodal editing, world knowledgeSafety featuresSynthID watermark, C2PA credentialsSynthID watermark, C2PA credentialsFrequently Asked QuestionsQ1. Who is Google positioning Nano Banana 2 Lite for? Google says Nano Banana 2 Lite is designed primarily for developers and enterprise users who need to generate and edit images quickly for applications such as social media, marketing campaigns, storyboarding and large-scale creative workflows.Q2. What improvements does Nano Banana 2 Lite bring over the previous Nano Banana model? The new model offers better character consistency, improved contextual understanding, enhanced text rendering and localisation, making it more suitable for complex image generation and editing tasks.Q3. How is Gemini Omni Flash different from Nano Banana 2 Lite? While Nano Banana 2 Lite focuses on AI-powered image generation and editing, Gemini Omni Flash is designed for conversational video generation and editing using text, image and video inputs.Q4. What measures has Google introduced to identify AI-generated content? Google says both Nano Banana 2 Lite and Gemini Omni Flash include C2PA content credentials and SynthID invisible watermarks by default to help improve transparency and verify AI-generated content.end of article