Gemini 2.0 Flash Multimodal Video & Audio Real-Time Analyzer
Master instructions for processing multi-hour video feeds, raw audio waveforms, and complex infographics with Google Gemini 2.0 Flash API with sub-second latency.
Parametric Customizer & Compiler Live Tool
Adjust the variables below. The prompt code and API payload update dynamically in real time.
Specifies video analysis target.
Determines structured payload format.
Sets resolution of timeline parsing.
You are a Principal AI Vision & Multimodal Architect utilizing Google Gemini 2.0 Flash. OBJECTIVE: Analyze the input multimodal stream (video timestamps, audio speech cadence, and visual frames) with the following parameters: Analysis Objective: Audit 60-minute technical product demo for UI UX bugs and confusing UX friction points Extraction Schema: Strict JSON Array of Timestamped Objects with Bounding Boxes and Confidence Scores Timestamp Granularity: Sub-second millisecond accuracy (e.g. 00:14:22.450) Strict Rule: Never hallucinate objects not physically present in the frame bounding boxes. MULTIMODAL PARSING PIPELINE: 1. CHRONOLOGICAL EVENT INDEX: Map all scene changes, speaker shifts, and textual overlays with millisecond accuracy. 2. OCR & VISUAL RECOGNITION: Transcribe on-screen diagrams, UI elements, code snippets, and whiteboards. 3. SENTIMENT & PROSODY: Detect vocal hesitation, tone shifts, and acoustic background anomalies. 4. STRUCTURED EXPORT: Return strictly valid JSON adhering to the specified Zod schema without conversational filler.
Run Gemini 2.0 Flash with free tier API access in Google AI Studio.
Leonardo Phoenix & Flux: Generate photorealistic studio assets with 150 daily free credits. Zero setup required.
πΈ Real Sample Outputs & Verified Results
Visual evidence and benchmark responses generated using this exact prompt formula.
π¬ Prompt Engineering Anatomy
Why this prompt consistently delivers high-tier results: technical parameter breakdown.
Gemini 2.0 Flash 1M Context Window Gemini 2.0 Flash natively accepts up to 1 million tokens of raw video frames (at 1 frame per second) and audio channels in a single API call.
Zero-Conversational Filler Mandate Prevents Gemini from prepending 'Here is your analysis' before the JSON payload, ensuring automated backend ingestion never breaks.
π¬ Frequently Asked Questions
Troubleshooting common edge cases, model incompatibilities, and best practices.
Can I use this with Google AI Studio or Vertex AI?
Yes! This prompt is optimized for Gemini 2.0 Flash via the `@google/genai` Node.js SDK and Google Cloud Vertex AI.
Does Gemini 2.0 Flash support audio understanding?
Yes! Gemini 2.0 natively analyzes audio tone, inflection, and background acoustics without requiring a separate Whisper speech-to-text pass.
Cursor AI & Claude 3.5 Sonnet Pro Bundle
Supercharge your workflow with instant codebase indexing, multi-file edits, and automated system rules.
Explore Related AI Coding & Architecture Prompts View all β
Claude 3.5 Sonnet Senior Staff Full-Stack Refactoring Architect
Specialized prompt framework for Claude 3.5 Sonnet and Claude 3.7. Transforms messy legacy codebases into modular, testable, high-performance systems with automated regression tests.
cursorCursor AI (.cursorrules) for Next.js 15, React 19 & Tailwind CSS
Stops Cursor from writing outdated Next.js 13 patterns, Pages Router syntax, or deprecated React hooks. Enforces strict TypeScript, Server Actions, and modular clean architecture.
deepseekDeepSeek-R1 Autonomous Deep Research & Synthesis Engine
DeepSeek-R1 chain-of-thought framework designed to force deep reflection, counterfactual auditing, and hypothesis testing before producing production-ready research reports.