[{"data":1,"prerenderedAt":30},["ShallowReactive",2],{"nr-en-google-gemini-agentic-video-understanding":3},{"slug":4,"title":5,"dek":6,"date":7,"time":8,"publishedAt":9,"updated":10,"updatedAt":10,"dateFmt":11,"updatedFmt":10,"kind":12,"tier":13,"author":14,"authorName":15,"topics":16,"tracker":22,"trackerLabel":23,"headlineStat":24,"image":25,"ogImage":26,"imageAlt":5,"csv":10,"minutes":27,"words":28,"html":29},"google-gemini-agentic-video-understanding","Google Makes Video Analysis Smarter: Gemini Cuts Token Use by 88%","With 'agentic video understanding,' Gemini's latest models can intelligently analyze videos instead of processing them frame-by-frame. The result: massive cost savings and better accuracy.","2026-09-02","09:59","2026-09-02T09:59:00+02:00","","September 2, 2026","news","standard","ideal-syka","Ideal Syka",[17,18,19,20,21],"Gemini","Video Analysis","AI Costs","Google DeepMind","Agentic AI","\u002Fki-preis","AI Pricing & Token Costs","88% token reduction for video analysis","\u002Fnewsroom\u002Fimg\u002Fgoogle-gemini-agentic-video-understanding.webp","\u002Fog-nr\u002Fgoogle-gemini-agentic-video-understanding.en.png",3,511,"\u003Cp>Google has launched a new capability for its Gemini models that fundamentally changes how video analysis works: \u003Cstrong>agentic video understanding\u003C\u002Fstrong>. Instead of processing videos frame-by-frame at a fixed rate, the models can now actively decide which segments to watch, how fast to scan them, and whether to use frames, audio, or transcripts. The results are impressive – and most importantly: significantly cheaper.\u003C\u002Fp>\n\u003Ch2>Quick Facts\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cstrong>Available for\u003C\u002Fstrong>: Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite via Gemini API, Google AI Studio, and Gemini Enterprise Agent Platform\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Savings\u003C\u002Fstrong>: Token consumption reduced by up to \u003Cstrong>88%\u003C\u002Fstrong>, costs cut by up to \u003Cstrong>66%\u003C\u002Fstrong>, accuracy improved by up to \u003Cstrong>7%\u003C\u002Fstrong>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Most effective for\u003C\u002Fstrong>: Long-form videos (10-minute tutorials to 90-minute lectures and multi-hour recordings)\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Launch\u003C\u002Fstrong>: Available now for video uploads and YouTube videos\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>How the Old System Worked – and Why It Was Inefficient\u003C\u002Fh2>\n\u003Cp>Previously, Gemini processed videos in \u003Cstrong>static mode\u003C\u002Fstrong>: the model ingested the entire file at a fixed frame rate (default 1 FPS, adjustable). For long videos, this created a dilemma: either high token costs or quality loss through sampling. Especially with long-form content (lectures, tutorials, multi-hour recordings), critical details were lost or the bill became unaffordable.\u003C\u002Fp>\n\u003Ch2>The New Approach: Targeted Video Analysis\u003C\u002Fh2>\n\u003Cp>Agentic video understanding works differently. The model takes \u003Cstrong>active control\u003C\u002Fstrong>: it decides which video segments are relevant, scans them at variable speeds, leverages parallel data sources (frames, audio, transcripts), and jumps directly to interesting moments. This is similar to \u003Cstrong>agentic vision\u003C\u002Fstrong> for images – but for videos.\u003C\u002Fp>\n\u003Cp>The results are impressive benchmarks:\u003C\u002Fp>\n\u003Cdiv class=\"tbl-scroll\">\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Metric\u003C\u002Fth>\n\u003Cth>Improvement\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>Token consumption\u003C\u002Ftd>\n\u003Ctd>up to \u003Cstrong>88% reduction\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Costs\u003C\u002Ftd>\n\u003Ctd>up to \u003Cstrong>66% savings\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Accuracy\u003C\u002Ftd>\n\u003Ctd>up to \u003Cstrong>7% improvement\u003C\u002Fstrong>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\u003C\u002Fdiv>\n\u003Cp>Particularly noteworthy: \u003Cstrong>Gemini 3.7 Flash\u003C\u002Fstrong> with agentic understanding now sits at the "accuracy-to-cost Pareto frontier" – meaning no other tested model offers better accuracy at the same cost or lower cost at the same accuracy.\u003C\u002Fp>\n\u003Ch2>New Use Cases Become Possible\u003C\u002Fh2>\n\u003Cp>The capability enables applications that were previously too expensive or too inaccurate: \u003Cstrong>sub-second moment retrieval\u003C\u002Fstrong> ("Find the exact moment the product is shown"), more precise \u003Cstrong>anomaly detection\u003C\u002Fstrong> ("When does something unusual happen?"), accurate \u003Cstrong>counting\u003C\u002Fstrong> ("How many people are in the video?"). For enterprises automating video analysis, this is a genuine capability upgrade.\u003C\u002Fp>\n\u003Cp>Activation is simple via API – just set the configuration to "agentic" in Google AI Studio or the Gemini Enterprise Agent Platform. Works with video uploads and YouTube videos.\u003C\u002Fp>\n\u003Ch2>What This Means for Enterprises\u003C\u002Fh2>\n\u003Cp>For companies integrating video into their KI workflows – from media organizations to security to document analysis – the cost equation changes significantly. 66% cost savings with better quality makes video-KI applications suddenly more economical. Especially for longer content (training videos, surveillance footage, archive analysis), this could be a tipping point. However: the capability is currently limited to Google's models – organizations already invested in other platforms need to re-evaluate whether switching makes sense.\u003C\u002Fp>\n\u003Ch2>Sources\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fmodels-and-research\u002Fgemini-models\u002Fintroducing-agentic-video-in-gemini\u002F\">blog.google\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cem>Editorially owned by \u003Ca href=\"\u002Fen\u002Fautor\u002Fideal-syka\">Ideal Syka\u003C\u002Fa>. Sources and method: \u003Ca href=\"\u002Fen\u002Fredaktion\">Newsroom & method\u003C\u002Fa>. Tips and corrections: \u003Ca href=\"mailto:ai@i6eal.de\">ai@i6eal.de\u003C\u002Fa>.\u003C\u002Fem>\u003C\u002Fp>\n",1788341041893]