Multimodal & Image Generation
Image, audio and video input and image generation.
Articles in Multimodal & Image Generation
- Gemini Video Understanding: Analyzing Video Content with Timestamp References — Gemini can watch a video, not just generate one -- sampling it at 1 frame per second, reasoning over audio and visuals together, and answering questions tied to a specific timestamp. Here is how native video analysis works.
- Gemini PDF and Document Understanding: Native Parsing for Tables, Charts, and Layout — Gemini reads PDFs the way a person does -- as pages with layout, tables, and charts -- not as extracted plain text. Here is how native document understanding works and what it costs in tokens.
- Gemini Files API Guide: Uploading Large Media for Multimodal Prompts — Anything over 100MB inline has to go through the Files API — but uploaded files vanish after 48 hours, which breaks pipelines that assume they persist. Limits, expiry, and Python code.
- Nano Banana 2 API Guide: Production-Scale Image Generation — Nano Banana 2 brings Gemini 3 intelligence to production-scale image generation with resolution-tiered pricing from 0.5K to 4K. Full API guide with pricing and Python/JS code.
- Nano Banana Pro: Studio-Grade 4K Visuals with Accurate Text — Nano Banana Pro is Gemini's studio-quality image model, built for 4K output and accurate in-image text rendering. Full pricing, code, and guidance on when it's worth the premium price.
- Imagen 4 Deprecation Guide: Migrating to Gemini's Nano Banana Models — Imagen 4 is marked deprecated in Google's current Gemini API docs. This guide covers why it's being replaced, which Nano Banana model to migrate to, and complete before/after code.
- Nano Banana 2 Lite Explained: Fast, Cheap Gemini Image Generation — Nano Banana 2 Lite trades peak fidelity for ultra-low latency and cost, making it the right choice for high-volume image generation. Full pricing breakdown and Python/JS code included.
- Nano Banana API Guide: Native Image Generation with Gemini — Nano Banana is Gemini's original native image generation and editing model. Here's the exact model ID, API method, pricing, and complete Python and JavaScript code to generate and edit images with it.
- Gemini 3.6 Flash API Guide: Multimodal Agentic Tasks with the Gemini API — Gemini 3.6 Flash is Google's generalist multimodal agentic model. Model ID, verified pricing, and working Python/JS examples for document and image reasoning tasks.