[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"nav-categories":3,"article-deepseek-ocr-a-new-approach-to-memory-in-ai":70},{"data":4},[5,37,57,64],{"name":6,"slug":7,"categories":8},"Productivity","productivity",[9,13,17,21,25,29,33],{"id":10,"title":11,"slug":12},17,"Branding","branding",{"id":14,"title":15,"slug":16},19,"Marketing","marketing",{"id":18,"title":19,"slug":20},20,"Work","work",{"id":22,"title":23,"slug":24},34,"Community","community",{"id":26,"title":27,"slug":28},21,"For newbies","for-newbies",{"id":30,"title":31,"slug":32},24,"Investment","investment",{"id":34,"title":35,"slug":36},22,"Finance","finance",{"name":38,"slug":39,"categories":40},"Tech","tech",[41,45,49,53],{"id":42,"title":43,"slug":44},28,"Technology","technology",{"id":46,"title":47,"slug":48},32,"Artificial Intelligence","artificial-intelligence",{"id":50,"title":51,"slug":52},26,"Security and protection","security-and-protection",{"id":54,"title":55,"slug":56},31,"YouTube Blog","youtube-blog",{"name":58,"slug":59,"categories":60},"News","news",[61],{"id":62,"title":58,"slug":63},18,"quasanews",{"name":65,"slug":66,"categories":67},"Business","business",[68],{"id":69,"title":65,"slug":66},16,{"post":71,"published_news":95,"popular_news":154,"categories":219},{"title":72,"description":73,"meta_title":72,"meta_description":74,"meta_keywords":75,"text":76,"slug":77,"created_at":78,"publish_at":79,"formatted_created_at":80,"category_id":42,"links":81,"view_type":84,"video_url":85,"views":86,"likes":87,"lang":88,"comments_count":87,"category":89},"DeepSeek-OCR: A New Approach to Memory in AI","In a groundbreaking move, DeepSeek has demonstrated that an Optical Character Recognition (OCR) model can do more than just read documents—it can revolutionize how AI models handle memory. The concept is both simple and audacious: rather than storing context in traditional text tokens, DeepSeek proposes encoding dialogue history as images of pages, which are then retrieved and processed via OCR as needed. This innovative approach could redefine long-term memory in large language models (LLMs), offering a fresh perspective on efficiency and retention.","For agents and LLMs tasked with long-term tasks, DeepSeek-OCR might just be the memory upgrade we’ve been waiting for.","This innovative approach could redefine long-term memory in large language models (LLMs), offering a fresh perspective on efficiency and retention","\u003Cp>In a groundbreaking move, DeepSeek has demonstrated that an Optical Character Recognition (OCR) model can do more than just read documents - it can revolutionize how AI models handle memory.\u003C/p>\n\n\u003Cp>The concept is both simple and audacious: rather than storing context in traditional text tokens, DeepSeek proposes encoding dialogue history as images of pages, which are then retrieved and processed via OCR as needed. This innovative approach could redefine long-term memory in large language models (LLMs), offering a fresh perspective on efficiency and retention.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>The Core Idea\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Cimg alt=\"\" class=\"image-align-left\" height=\"289\" src=\"https://quasa.io/storage/photos/00/photo_2025-11-02_18-30-31.jpg\" width=\"300\" />The brilliance of DeepSeek&rsquo;s method lies in its shift from text-based tokens to visual representations. Typically, LLMs rely on tokenizing text, where each token represents a fragment of language. However, a single visual patch - an image snippet - can encapsulate far more information than a single text token. By rendering conversation history as pages and storing them as images, the model maintains a compact visual representation. Precise citations or details are extracted only when requested through OCR, striking a balance between memory retention and computational efficiency.\u003C/p>\n\n\u003Cp>This approach allows the model to preserve more context using fewer resources. Instead of burning through thousands of tokens to maintain a lengthy dialogue, DeepSeek&rsquo;s system leverages hundreds of visual patches, significantly reducing the token overhead and cutting costs associated with long-context processing.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>What Happens Under the Hood\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Cstrong>\u003Cimg alt=\"\" class=\"image-align-right\" height=\"169\" src=\"https://quasa.io/storage/photos/00/image - 2025-11-02T183822.214.jpg\" width=\"300\" />The mechanics of this system are elegantly designed:\u003C/strong>\u003C/p>\n\n\u003Cul>\n\t\u003Cli>\u003Cstrong>History Packaging:\u003C/strong>&nbsp;The dialogue history is organized into pages and segmented into 2D patches.\u003C/li>\n\t\u003Cli>\u003Cstrong>Quality Tiering:\u003C/strong>&nbsp;Recent pages are stored in high resolution to ensure clarity, while older pages are compressed more aggressively but not discarded entirely.\u003C/li>\n\t\u003Cli>\u003Cstrong>On-Demand OCR:\u003C/strong>&nbsp;The OCR module is triggered only when a specific word, line, or detail is required, avoiding unnecessary processing.\u003C/li>\n\u003C/ul>\n\n\u003Cp>This creates a form of &quot;soft memory fading&quot; rather than the abrupt context truncation common in traditional models. Crucially, structured elements like tables, code, and text formatting remain intact, enabling the model to retain contextual anchors that might otherwise be lost.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Practical Impact\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Cstrong>\u003Cimg alt=\"\" class=\"image-align-right\" height=\"169\" src=\"https://quasa.io/storage/photos/00/image - 2025-11-02T184032.499.jpg\" width=\"300\" />The implications of this method are substantial:\u003C/strong>\u003C/p>\n\n\u003Cul>\n\t\u003Cli>\u003Cstrong>Token Efficiency:\u003C/strong>&nbsp;Thousands of text tokens are condensed into hundreds of visual patches, slashing the resource demand.\u003C/li>\n\t\u003Cli>\u003Cstrong>Cost-Effectiveness:\u003C/strong>&nbsp;Lower token usage translates to reduced computational costs, making it viable for broader deployment.\u003C/li>\n\t\u003Cli>\u003Cstrong>Agent-Friendly Design:\u003C/strong>&nbsp;This approach is particularly suited for agent-based systems that manage extended sessions, revisit past actions, or analyze logs over time.\u003C/li>\n\t\u003Cli>\u003Cstrong>Self-Generated Training Data:\u003C/strong>&nbsp;The system can render pages and generate OCR labels on the fly, creating a self-sustaining loop for training data production.\u003C/li>\n\u003C/ul>\n\n\u003Cp>While this method doesn&rsquo;t make the model infallible at memorization, it extends the duration of information retention and enables selective recall without relying on external storage or complex Retrieval-Augmented Generation (RAG) frameworks.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>A Paradigm Shift for Long-Term Memory\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Cimg alt=\"\" class=\"image-align-left\" height=\"169\" src=\"https://quasa.io/storage/photos/00/image - 2025-11-02T183940.386.jpg\" width=\"300\" />Storing text as images and reading it on demand could emerge as a new paradigm for long-term memory in LLMs. This is especially promising for AI agents that need to track a journey - remembering the full path rather than just the latest step. Traditional models often struggle with context windows that limit how far back they can look, but DeepSeek-OCR offers a scalable alternative. By compressing historical data into visual snapshots and retrieving specifics as needed, the system mimics human-like memory decay while preserving accessibility.\u003C/p>\n\n\u003Chr />\n\u003Cp>\u003Cstrong>\u003Cimg alt=\"\" class=\"image-align-right\" height=\"169\" src=\"https://quasa.io/storage/photos/00/image - 2025-11-02T183954.910.jpg\" width=\"300\" />Also read:\u003C/strong>\u003C/p>\n\n\u003Cul>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/how-youtuber-and-indie-hacker-marc-lou-tackled-the-fake-revenue-screenshot-problem\">How YouTuber and Indie Hacker Marc Lou Tackled the Fake Revenue Screenshot Problem\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/openai-is-turning-chatgpt-into-a-boring-nerd\">OpenAI Is Turning ChatGPT into a Boring Nerd\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/emerging-technologies-and-approaches-in-digital-marketing-for-2026\">Emerging Technologies and Approaches in Digital Marketing for 2026\u003C/a>\u003C/li>\n\u003C/ul>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>The Future of AI Memory\u003C/strong>\u003C/h4>\n\n\u003Cp>As of 06:30 PM CET on November 07, 2025, DeepSeek&rsquo;s OCR-based memory approach is still in its early stages, but the potential is clear. This technique could pave the way for more robust, efficient, and context-aware AI systems, particularly in applications requiring prolonged engagement or detailed historical analysis. While it may not replace all existing memory mechanisms, it complements them by offering a lightweight, innovative solution. For agents and LLMs tasked with long-term tasks, DeepSeek-OCR might just be the memory upgrade we&rsquo;ve been waiting for.\u003C/p>","deepseek-ocr-a-new-approach-to-memory-in-ai","2025-11-02T17:42:46.000000Z","2025-11-07T09:32:00.000000Z","07.11.2025",{"image":82,"thumb":83},"https://quasa.io/storage/images/news/QcaPdMXp2zgAKx3cNlMig9QxbuPkYbBoXQIyDoJh.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/QcaPdMXp2zgAKx3cNlMig9QxbuPkYbBoXQIyDoJh.jpg","small",null,1338,0,"en",{"id":42,"title":43,"slug":44,"meta_title":90,"meta_description":91,"meta_keywords":92,"deleted_at":85,"created_at":93,"updated_at":94,"lang":88},"Quasa media blog about growth hacking in Tech","All the most interesting and useful about technologies. Exclusive articles from technologies you won't find anywhere else.","Technology, tech, business, ai, gadget, gadgets, life hacks","2023-03-23T08:15:32.000000Z","2024-08-25T15:37:57.000000Z",[96,109,120,131,142],{"title":97,"description":98,"slug":99,"created_at":100,"publish_at":101,"formatted_created_at":102,"category":103,"links":104,"view_type":84,"video_url":85,"views":107,"likes":87,"lang":88,"comments_count":87,"is_pinned":108},"China’s Five-Year Plans Strike Again: How Centralized Vision and Competitive Freedom Are Powering the Next Frontier of Brain-Computer Interfaces","In an era of breakneck technological change, China’s much-maligned five-year planning system is proving surprisingly effective. Far from the rigid, top-down micromanagement of the Soviet era, Beijing’s modern industrial strategies deliberately avoid over-specifying every detail.","china-s-five-year-plans-strike-again-how-centralized-vision-and-competitive-freedom-are-powering-the-next-frontier-of-brain-computer-interfaces","2026-03-28T17:45:42.000000Z","2026-04-10T11:36:00.000000Z","10.04.2026",{"title":43,"slug":44},{"image":105,"thumb":106},"https://quasa.io/storage/images/news/9e878UicRgHXBtTQ74llERUUJHi9VJhY6RrS6GzZ.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/9e878UicRgHXBtTQ74llERUUJHi9VJhY6RrS6GzZ.jpg",70,false,{"title":110,"description":111,"slug":112,"created_at":113,"publish_at":114,"formatted_created_at":102,"category":115,"links":116,"view_type":84,"video_url":85,"views":119,"likes":87,"lang":88,"comments_count":87,"is_pinned":108},"This AI Will Tell You Exactly How Attractive You Are — And It Only Takes 25 Seconds","There’s a new viral AI tool that does something most of us secretly want to know but are afraid to ask: it looks at your face and gives you a straight-up attractiveness score.","this-ai-will-tell-you-exactly-how-attractive-you-are-and-it-only-takes-25-seconds","2026-03-27T20:13:20.000000Z","2026-04-10T09:34:00.000000Z",{"title":27,"slug":28},{"image":117,"thumb":118},"https://quasa.io/storage/images/news/eTdAX16TIQnnM1X90hkB2oOPixcWqRw3eZlaGUta.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/eTdAX16TIQnnM1X90hkB2oOPixcWqRw3eZlaGUta.jpg",200,{"title":121,"description":122,"slug":123,"created_at":124,"publish_at":125,"formatted_created_at":102,"category":126,"links":127,"view_type":84,"video_url":85,"views":130,"likes":87,"lang":88,"comments_count":87,"is_pinned":108},"Nubia Unveils RedMagic 11 Pro: The World’s First Gaming Smartphone with Liquid Cooling","The new flagship features a ceramic micropump, a massive 8,000mAh battery, and a Snapdragon 8 Elite Gen 5 processor.","nubia-unveils-redmagic-11-pro-the-world-s-first-gaming-smartphone-with-liquid-cooling","2026-03-26T06:45:27.000000Z","2026-04-10T06:40:00.000000Z",{"title":43,"slug":44},{"image":128,"thumb":129},"https://quasa.io/storage/images/news/g3jNseMCCzEO2cZreMfoeEH3R0kh2VsnDf511jwh.png","https://api.quasa.io/thumbs/news-thumb/images/news/g3jNseMCCzEO2cZreMfoeEH3R0kh2VsnDf511jwh.png",240,{"title":132,"description":133,"slug":134,"created_at":135,"publish_at":136,"formatted_created_at":102,"category":137,"links":138,"view_type":84,"video_url":85,"views":141,"likes":87,"lang":88,"comments_count":87,"is_pinned":108},"Chinese Startup Hypershell Unveils X Ultra Exoskeleton — More Powerful Than an E-Bike","Hypershell, a Chinese company specializing in wearable robotics, has just launched one of the most impressive consumer exoskeletons to date: the Hypershell X Ultra.","chinese-startup-hypershell-unveils-x-ultra-exoskeleton-more-powerful-than-an-e-bike","2026-03-25T21:36:46.000000Z","2026-04-10T03:27:00.000000Z",{"title":43,"slug":44},{"image":139,"thumb":140},"https://quasa.io/storage/images/news/xgiykS2Xj3WKWW2BoMKNqCWFJgIrJ1rfykipZMXy.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/xgiykS2Xj3WKWW2BoMKNqCWFJgIrJ1rfykipZMXy.jpg",338,{"title":143,"description":144,"slug":145,"created_at":146,"publish_at":147,"formatted_created_at":148,"category":149,"links":150,"view_type":84,"video_url":85,"views":153,"likes":87,"lang":88,"comments_count":87,"is_pinned":108},"Your Computer Mouse Can Now Eavesdrop on You","Researchers from the University of California, Irvine have demonstrated a surprising new security threat: an ordinary optical computer mouse can be turned into a covert microphone capable of listening to your conversations.","your-computer-mouse-can-now-eavesdrop-on-you","2026-03-25T21:21:22.000000Z","2026-04-09T11:13:00.000000Z","09.04.2026",{"title":43,"slug":44},{"image":151,"thumb":152},"https://quasa.io/storage/images/news/zvGSPNFVpmuJgnlniEHiwPc4aaeGYIVAuwxhouXu.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/zvGSPNFVpmuJgnlniEHiwPc4aaeGYIVAuwxhouXu.jpg",896,[155,168,182,194,207],{"title":156,"description":157,"slug":158,"created_at":159,"publish_at":160,"formatted_created_at":161,"category":162,"links":163,"view_type":84,"video_url":85,"views":166,"likes":167,"lang":88,"comments_count":87,"is_pinned":108},"The Anatomy of an Entrepreneur","Entrepreneur is a French word that means an enterpriser. Enterprisers are people who undertake a business or enterprise with the chance of earning profits or suffering from loss.","the-anatomy-of-an-entrepreneur","2021-08-04T15:18:21.000000Z","2025-12-14T06:09:00.000000Z","14.12.2025",{"title":65,"slug":66},{"image":164,"thumb":165},"https://quasa.io/storage/images/news/mVsXPTMuHZuI7UXCsENgL1Qwp1uSOf7Rz3uVPMfm.webp","https://api.quasa.io/thumbs/news-thumb/images/news/mVsXPTMuHZuI7UXCsENgL1Qwp1uSOf7Rz3uVPMfm.webp",68830,2,{"title":169,"description":170,"slug":171,"created_at":172,"publish_at":173,"formatted_created_at":174,"category":175,"links":176,"view_type":179,"video_url":85,"views":180,"likes":181,"lang":88,"comments_count":87,"is_pinned":108},"Advertising on QUASA","QUASA MEDIA is read by more than 400 thousand people a month. We offer to place your article, add a link or order the writing of an article for publication.","advertising-on-quasa","2022-07-06T07:33:02.000000Z","2025-12-15T17:33:02.000000Z","15.12.2025",{"title":58,"slug":63},{"image":177,"thumb":178},"https://quasa.io/storage/images/news/45SvmdsTQbiyc3nxgbyHY1mpVbisYyub2BCHjqBL.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/45SvmdsTQbiyc3nxgbyHY1mpVbisYyub2BCHjqBL.jpg","large",68409,4,{"title":183,"description":184,"slug":185,"created_at":186,"publish_at":187,"formatted_created_at":188,"category":189,"links":190,"view_type":84,"video_url":85,"views":193,"likes":181,"lang":88,"comments_count":87,"is_pinned":108},"What is a Startup?","A startup is not a new company, not a tech company, nor a new tech company. You can be a new tech company, if your goal is not to grow high and fast; then, you are not a startup. ","what-is-a-startup","2021-08-04T12:05:17.000000Z","2025-12-17T13:02:00.000000Z","17.12.2025",{"title":65,"slug":66},{"image":191,"thumb":192},"https://quasa.io/storage/images/news/EOsQhSW3VXyG7a6NPdE1oZd00xfJXe3bjY5aJGb7.webp","https://api.quasa.io/thumbs/news-thumb/images/news/EOsQhSW3VXyG7a6NPdE1oZd00xfJXe3bjY5aJGb7.webp",66405,{"title":195,"description":196,"slug":197,"created_at":198,"publish_at":199,"formatted_created_at":200,"category":201,"links":202,"view_type":84,"video_url":85,"views":205,"likes":167,"lang":88,"comments_count":206,"is_pinned":108},"Top 5 Tips to Make More Money as a Content Creator","Content creators are one of the most desired job titles right now. Who wouldn’t want to earn a living online?","top-5-tips-to-make-more-money-as-a-content-creator","2022-01-17T17:31:51.000000Z","2026-01-17T11:30:00.000000Z","17.01.2026",{"title":19,"slug":20},{"image":203,"thumb":204},"https://quasa.io/storage/images/news/gP8kiumBPpJmQv6SMieXiX1tDetx43VwFfO1P4Ca.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/gP8kiumBPpJmQv6SMieXiX1tDetx43VwFfO1P4Ca.jpg",40666,1,{"title":208,"description":209,"slug":210,"created_at":211,"publish_at":212,"formatted_created_at":213,"category":214,"links":215,"view_type":179,"video_url":85,"views":218,"likes":167,"lang":88,"comments_count":87,"is_pinned":108},"8 Logo Design Tips for Small Businesses","Your logo tells the story of your business and the values you stand for.","8-logo-design-tips-for-small-businesses","2021-12-04T21:59:52.000000Z","2025-05-05T03:30:00.000000Z","05.05.2025",{"title":15,"slug":16},{"image":216,"thumb":217},"https://quasa.io/storage/images/news/Wbx2NtS1CnTupgoQbpFMGspJ5jm4uob2hDOq33r0.jpg","https://api.quasa.io/thumbs/news-thumb/images/news/Wbx2NtS1CnTupgoQbpFMGspJ5jm4uob2hDOq33r0.jpg",39960,[220,221,222,223,224,225,226,227,228,229,230,231,232],{"title":23,"slug":24},{"title":47,"slug":48},{"title":55,"slug":56},{"title":43,"slug":44},{"title":51,"slug":52},{"title":31,"slug":32},{"title":35,"slug":36},{"title":27,"slug":28},{"title":19,"slug":20},{"title":15,"slug":16},{"title":58,"slug":63},{"title":11,"slug":12},{"title":65,"slug":66}]