[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"nav-categories":3,"article-mistral-ai-teaches-small-models-to-think-like-giants-using-cascade-distillation":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":96,"popular_news":161,"categories":232},{"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":46,"links":81,"view_type":86,"video_url":87,"views":88,"likes":89,"lang":90,"comments_count":89,"category":91},"Mistral AI Teaches Small Models to Think Like Giants Using Cascade Distillation","In a significant advancement for efficient AI deployment, Mistral AI has unveiled the Ministral 3 family of models, leveraging an innovative technique called cascade distillation to create compact yet powerful vision-language models.","Mistral AI's cascade distillation represents a smart evolution in model development, empowering small models to perform like giants.","The Ministral 3 family not only achieves high precision with fewer parameters but also paves the way for more sustainable and scalable AI.","\u003Cp>In a significant advancement for efficient AI deployment, \u003Ca href=\"https://quasa.io/media/mistral-enhances-le-chat-with-deep-research-mode-and-advanced-features\">Mistral AI\u003C/a> has unveiled the \u003Ca href=\"https://www.deeplearning.ai/the-batch/mistral-uses-cascade-distillation-on-mistral-3-to-build-ministral-family\">Ministral 3 family of models, leveraging an innovative technique called cascade distillation to create compact yet powerful vision-language models\u003C/a>. This approach allows smaller models to inherit the "thinking" capabilities of larger ones, enabling high performance in resource-constrained environments like edge devices and local setups.\u003C/p>\n\n\u003Cp>By distilling knowledge from a robust "teacher" model through multiple stages, Mistral is bridging the gap between heavyweight AI training and lightweight production inference.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Understanding Cascade Distillation\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Cpicture class=\"image-align-right\">\u003Csource srcset=\"https://cdn.quasa.io/photos/00/image-2026-03-03t172809568.webp\" type=\"image/webp\">\u003Cimg alt=\"Mistral AI Teaches Small Models to Think Like Giants Using Cascade Distillation\" class=\"image-align-right\" height=\"447\" src=\"https://cdn.quasa.io/photos/00/image-2026-03-03t172809568.jpg\" width=\"300\" />\u003C/picture>Cascade distillation combines model pruning and knowledge distillation in a sequential manner, starting from a larger parent model. The process begins with Mistral Small 3.1, a 24-billion-parameter model acting as the primary "teacher."\u003C/p>\n\n\u003Cp>Pruning involves strategically removing layers that minimally impact inputs, reducing internal representation sizes, and narrowing fully connected layers. This creates an initial smaller model, such as the 14B variant, which then serves as the foundation for the next size down.\u003C/p>\n\n\u003Cp>Unlike traditional single-step distillation, the cascade method proceeds iteratively: each subsequent model learns from the outputs and refinements of the previous one. During pretraining, the pruned models are distilled by mimicking the teacher's outputs, with better results from Mistral Small 3.1 than the even larger Mistral Medium 3.\u003C/p>\n\n\u003Cp>Fine-tuning incorporates additional techniques like ODPO (Offline Direct Preference Optimization) for instruction-following and GRPO (Group Relative Policy Optimization) for reasoning variants, using step-by-step examples in math, coding, and more. Mistral Medium 3 assists in fine-tuning stages to enhance quality.\u003C/p>\n\n\u003Cp>This multi-stage process ensures that knowledge is transferred efficiently, resulting in models that "think" like their larger counterparts but with reduced computational demands.\u003C/p>\n\n\u003Cp>\u003Cstrong>Here is a visualization of the cascade distillation process:\u003C/strong>\u003C/p>\n\n\u003Cp>\u003Cstrong>\u003Cpicture>\u003Csource srcset=\"https://cdn.quasa.io/photos/00/2026-03-03-16-23-38-1.webp\" type=\"image/webp\">\u003Cimg alt=\"Mistral AI Teaches Small Models to Think Like Giants Using Cascade Distillation\" height=\"226\" src=\"https://cdn.quasa.io/photos/00/2026-03-03-16-23-38-1.jpg\" width=\"400\" />\u003C/picture>\u003C/strong>\u003C/p>\n\n\u003Ch4>\u003Cstrong>The Ministral 3 Family\u003C/strong>\u003C/h4>\n\n\u003Cp>The Ministral 3 lineup includes models with 14 billion, 8 billion, and 3 billion parameters, each available in base, instruction-tuned, and reasoning variants. These are open-weights vision-language models under the Apache 2.0 license, handling text and image inputs (up to 256,000 tokens for base, 128,000 for reasoning) and generating text outputs. They support multilingual capabilities in 11 languages, tool use, and a decoder-only transformer architecture.\u003C/p>\n\n\u003Cp>API pricing reflects their efficiency: $0.20 per million tokens for the 14B, $0.15 for the 8B, and $0.10 for the 3B. Training used just 1 to 3 trillion tokens—far fewer than competitors like Qwen 3 or Llama 3, which require 15 to 36 trillion—making the process quicker and less resource-intensive.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Performance Highlights\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Cpicture class=\"image-align-right\">\u003Csource srcset=\"https://cdn.quasa.io/photos/00/image-2026-03-03t172810925.webp\" type=\"image/webp\">\u003Cimg alt=\"Mistral AI Teaches Small Models to Think Like Giants Using Cascade Distillation\" class=\"image-align-right\" height=\"402\" src=\"https://cdn.quasa.io/photos/00/image-2026-03-03t172810925.jpg\" width=\"270\" />\u003C/picture>Despite their smaller sizes, Ministral models deliver impressive results, often rivaling or surpassing larger peers.\u003C/p>\n\n\u003Cp>The 14B Base matches or exceeds Mistral Small 3.1 on benchmarks like MATH (67.6% vs. competitors' lower scores), TriviaQA (74.9%), and GPQA Diamond. It ranks ahead of Mistral Small 3.1 and 3.2 on the Artificial Analysis Intelligence Index.\u003C/p>\n\n\u003Cp>\u003Cstrong>Comparisons show:\u003C/strong>\u003C/p>\n\n\u003Cul>\n\t\u003Cli>Ministral 3 14B outperforms Qwen 3 14B on MATH (67.6% vs. 62%) and TriviaQA (74.9% vs. 70.3%), though it trails Gemma 3 12B slightly on some.\u003C/li>\n\t\u003Cli>The 8B Base beats the larger Gemma 3 12B on most tests except TriviaQA.\u003C/li>\n\t\u003Cli>The 3B Base competes strongly with Gemma 3 4B and Qwen 3 4B, excelling on MATH.\u003C/li>\n\t\u003Cli>Reasoning variants shine on AIME 2025 (85% accuracy for 14B vs. 73.7% for Qwen 3 14B Thinking).\u003C/li>\n\u003C/ul>\n\n\u003Cp>These metrics demonstrate that cascade distillation preserves high accuracy while shrinking model size.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Benefits for Real-World Deployment\u003C/strong>\u003C/h4>\n\n\u003Cp>The Ministral family's key advantages lie in efficiency: faster inference times, lower production costs, and suitability for edge devices like laptops and smartphones. By using large models primarily for training and deploying small, distilled versions in production, organizations can scale AI services without prohibitive expenses. This approach reduces energy consumption and enables local, on-device AI, expanding accessibility for applications in mobile, IoT, and resource-limited settings.\u003C/p>\n\n\u003Cp>\u003Cpicture class=\"image-align-left\">\u003Csource srcset=\"https://cdn.quasa.io/photos/00/image-2026-03-03t172908608.webp\" type=\"image/webp\">\u003Cimg alt=\"Mistral AI Teaches Small Models to Think Like Giants Using Cascade Distillation\" class=\"image-align-left\" height=\"169\" src=\"https://cdn.quasa.io/photos/00/image-2026-03-03t172908608.jpg\" width=\"300\" />\u003C/picture>Also read:\u003C/p>\n\n\u003Cul>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/vibe-coding-and-icos-the-hamster-wheel-of-hype-for-the-average-user\">Vibe Coding and ICOs: The Hamster Wheel of Hype for the Average User\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/how-to-become-the-top-1-in-the-age-of-ai\">How to Become the Top 1% in the Age of AI\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/building-a-solid-foundation-for-marketing-strategy-development-using-chatgpt-gemini-and-claude\">Building a Solid Foundation for Marketing Strategy Development Using ChatGPT, Gemini, and Claude\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/partnership-announcement-quasa-token2049\">Partnership Announcement QUASA + TOKEN2049\u003C/a>\u003C/li>\n\u003C/ul>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Conclusion\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Ca href=\"https://quasa.io/projects/mistral-ai\">Mistral AI's cascade distillation represents a smart evolution in model development\u003C/a>, empowering small models to perform like giants. The Ministral 3 family not only achieves high precision with fewer parameters but also paves the way for more sustainable and scalable AI. As the field moves toward edge computing, techniques like this could democratize advanced AI, making powerful tools available beyond data centers. Developers and businesses can explore these models now, freely downloadable and ready to integrate.\u003C/p>","mistral-ai-teaches-small-models-to-think-like-giants-using-cascade-distillation","2026-03-03T16:30:52.000000Z","2026-03-08T11:22:00.000000Z","08.03.2026",{"image":82,"image_webp":83,"thumb":84,"thumb_webp":85},"https://cdn.quasa.io/images/news/ZqJjA3C2Rj2eWdKk2M1aaVqCj2jeJTMemucSs1e5.jpg","https://cdn.quasa.io/images/news/ZqJjA3C2Rj2eWdKk2M1aaVqCj2jeJTMemucSs1e5.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/ZqJjA3C2Rj2eWdKk2M1aaVqCj2jeJTMemucSs1e5.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/ZqJjA3C2Rj2eWdKk2M1aaVqCj2jeJTMemucSs1e5.webp","small",null,920,0,"en",{"id":46,"title":47,"slug":48,"meta_title":92,"meta_description":93,"meta_keywords":93,"deleted_at":87,"created_at":94,"updated_at":95,"lang":90},"Artificial Intelligence | AI Breakthroughs, Agents & Tools | QUASA","Artificial Intelligence, ai, ml, machine learning, chatgpt, future","2024-09-22T08:08:27.000000Z","2026-04-22T14:56:34.000000Z",[97,112,124,136,148],{"title":98,"description":99,"slug":100,"created_at":101,"publish_at":102,"formatted_created_at":103,"category":104,"links":105,"view_type":86,"video_url":87,"views":110,"likes":89,"lang":90,"comments_count":89,"is_pinned":111},"The Rice on the Chessboard: Why Humanity Keeps Underestimating Exponential AI Growth","There’s an ancient Indian parable about a wise man who, for a service rendered, asked a powerful king to pay him in rice — placing one grain on the first square of a chessboard, two on the second, four on the third, and doubling the amount with every subsequent square.","the-rice-on-the-chessboard-why-humanity-keeps-underestimating-exponential-ai-growth","2026-04-20T20:10:17.000000Z","2026-04-23T11:00:00.000000Z","23.04.2026",{"title":47,"slug":48},{"image":106,"image_webp":107,"thumb":108,"thumb_webp":109},"https://cdn.quasa.io/images/news/A7vLK20XbrRVdUCdvtedODguBAKoaZY79tMBlsOE.jpg","https://cdn.quasa.io/images/news/A7vLK20XbrRVdUCdvtedODguBAKoaZY79tMBlsOE.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/A7vLK20XbrRVdUCdvtedODguBAKoaZY79tMBlsOE.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/A7vLK20XbrRVdUCdvtedODguBAKoaZY79tMBlsOE.webp",8,false,{"title":113,"description":114,"slug":115,"created_at":116,"publish_at":117,"formatted_created_at":103,"category":118,"links":119,"view_type":86,"video_url":87,"views":62,"likes":89,"lang":90,"comments_count":89,"is_pinned":111},"AI Companies Are “Harvesting Organs” from Dead Startups — And Founders Are Cashing In","In the race for ever-better AI, the obvious data sources — the entire public internet, books, Reddit, Wikipedia — ran dry by late 2024.","ai-companies-are-harvesting-organs-from-dead-startups-and-founders-are-cashing-in","2026-04-19T18:19:11.000000Z","2026-04-23T09:10:00.000000Z",{"title":35,"slug":36},{"image":120,"image_webp":121,"thumb":122,"thumb_webp":123},"https://cdn.quasa.io/images/news/AuOWNag5KhQ2BXoZMgz0ynmxkQo4SwxeXqYAsges.jpg","https://cdn.quasa.io/images/news/AuOWNag5KhQ2BXoZMgz0ynmxkQo4SwxeXqYAsges.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/AuOWNag5KhQ2BXoZMgz0ynmxkQo4SwxeXqYAsges.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/AuOWNag5KhQ2BXoZMgz0ynmxkQo4SwxeXqYAsges.webp",{"title":125,"description":126,"slug":127,"created_at":128,"publish_at":128,"formatted_created_at":103,"category":129,"links":130,"view_type":86,"video_url":87,"views":135,"likes":89,"lang":90,"comments_count":89,"is_pinned":111},"10 Years in Crypto: QUASA Joins the Elite 1% and Continues to Gain Strength (QUA Crypto Buyback - April 2026)","As of now, the total QUA tokens in circulation stand at 67.95 million. This move is part of a broader strategy to optimize the token's economics amid favorable market conditions.","10-years-in-crypto-quasa-joins-the-elite-1-and-continues-to-gain-strength-qua-crypto-buyback-april-2026","2026-04-23T07:39:33.000000Z",{"title":58,"slug":63},{"image":131,"image_webp":132,"thumb":133,"thumb_webp":134},"https://cdn.quasa.io/images/news/2NQNDviIEG8GASJXJR2iLri59UokOqLzzlxHZ8zh.jpg","https://cdn.quasa.io/images/news/2NQNDviIEG8GASJXJR2iLri59UokOqLzzlxHZ8zh.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/2NQNDviIEG8GASJXJR2iLri59UokOqLzzlxHZ8zh.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/2NQNDviIEG8GASJXJR2iLri59UokOqLzzlxHZ8zh.webp",25,{"title":137,"description":138,"slug":139,"created_at":140,"publish_at":141,"formatted_created_at":103,"category":142,"links":143,"view_type":86,"video_url":87,"views":22,"likes":89,"lang":90,"comments_count":89,"is_pinned":111},"The Death of a VR Unicorn: Rec Room Shuts Down After 150 Million Users, $294 Million Raised, and a $3.5 Billion Valuation","After a decade of building one of the most beloved social VR platforms on the planet, Rec Room is closing its doors for good.","the-death-of-a-vr-unicorn-rec-room-shuts-down-after-150-million-users-294-million-raised-and-a-3-5-billion-valuation","2026-04-19T15:55:34.000000Z","2026-04-23T06:49:00.000000Z",{"title":65,"slug":66},{"image":144,"image_webp":145,"thumb":146,"thumb_webp":147},"https://cdn.quasa.io/images/news/AJj9CTvP0kh4rWQeTWBUNuEH7ZDRAAumuG3vyFxP.jpg","https://cdn.quasa.io/images/news/AJj9CTvP0kh4rWQeTWBUNuEH7ZDRAAumuG3vyFxP.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/AJj9CTvP0kh4rWQeTWBUNuEH7ZDRAAumuG3vyFxP.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/AJj9CTvP0kh4rWQeTWBUNuEH7ZDRAAumuG3vyFxP.webp",{"title":149,"description":150,"slug":151,"created_at":152,"publish_at":153,"formatted_created_at":103,"category":154,"links":155,"view_type":86,"video_url":87,"views":160,"likes":89,"lang":90,"comments_count":89,"is_pinned":111},"Ikea Keeps the Human Face in the Age of AI: How Chatbot Billie Created More Creative Jobs — and $1.3 Billion in New Revenue","In an industry where many companies are racing to replace customer-service staff with AI, Swedish furniture giant Ikea has taken the opposite path.","ikea-keeps-the-human-face-in-the-age-of-ai-how-chatbot-billie-created-more-creative-jobs-and-1-3-billion-in-new-revenue","2026-04-19T14:33:29.000000Z","2026-04-23T03:23:00.000000Z",{"title":65,"slug":66},{"image":156,"image_webp":157,"thumb":158,"thumb_webp":159},"https://cdn.quasa.io/images/news/kPnnP95Xkml8A4gkXzaPwD34FPO0oXCjf1rcxWxU.jpg","https://cdn.quasa.io/images/news/kPnnP95Xkml8A4gkXzaPwD34FPO0oXCjf1rcxWxU.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/kPnnP95Xkml8A4gkXzaPwD34FPO0oXCjf1rcxWxU.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/kPnnP95Xkml8A4gkXzaPwD34FPO0oXCjf1rcxWxU.webp",51,[162,175,191,203,218],{"title":163,"description":164,"slug":165,"created_at":166,"publish_at":167,"formatted_created_at":168,"category":169,"links":170,"view_type":86,"video_url":87,"views":173,"likes":174,"lang":90,"comments_count":89,"is_pinned":111},"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":171,"image_webp":87,"thumb":172,"thumb_webp":172},"https://cdn.quasa.io/images/news/mVsXPTMuHZuI7UXCsENgL1Qwp1uSOf7Rz3uVPMfm.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/mVsXPTMuHZuI7UXCsENgL1Qwp1uSOf7Rz3uVPMfm.webp",71305,2,{"title":176,"description":177,"slug":178,"created_at":179,"publish_at":180,"formatted_created_at":181,"category":182,"links":183,"view_type":188,"video_url":87,"views":189,"likes":190,"lang":90,"comments_count":89,"is_pinned":111},"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":184,"image_webp":185,"thumb":186,"thumb_webp":187},"https://cdn.quasa.io/images/news/45SvmdsTQbiyc3nxgbyHY1mpVbisYyub2BCHjqBL.jpg","https://cdn.quasa.io/images/news/45SvmdsTQbiyc3nxgbyHY1mpVbisYyub2BCHjqBL.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/45SvmdsTQbiyc3nxgbyHY1mpVbisYyub2BCHjqBL.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/45SvmdsTQbiyc3nxgbyHY1mpVbisYyub2BCHjqBL.webp","large",71078,4,{"title":192,"description":193,"slug":194,"created_at":195,"publish_at":196,"formatted_created_at":197,"category":198,"links":199,"view_type":86,"video_url":87,"views":202,"likes":190,"lang":90,"comments_count":89,"is_pinned":111},"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":200,"image_webp":87,"thumb":201,"thumb_webp":201},"https://cdn.quasa.io/images/news/EOsQhSW3VXyG7a6NPdE1oZd00xfJXe3bjY5aJGb7.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/EOsQhSW3VXyG7a6NPdE1oZd00xfJXe3bjY5aJGb7.webp",68696,{"title":204,"description":205,"slug":206,"created_at":207,"publish_at":208,"formatted_created_at":209,"category":210,"links":211,"view_type":86,"video_url":87,"views":216,"likes":174,"lang":90,"comments_count":217,"is_pinned":111},"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":212,"image_webp":213,"thumb":214,"thumb_webp":215},"https://cdn.quasa.io/images/news/gP8kiumBPpJmQv6SMieXiX1tDetx43VwFfO1P4Ca.jpg","https://cdn.quasa.io/images/news/gP8kiumBPpJmQv6SMieXiX1tDetx43VwFfO1P4Ca.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/gP8kiumBPpJmQv6SMieXiX1tDetx43VwFfO1P4Ca.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/gP8kiumBPpJmQv6SMieXiX1tDetx43VwFfO1P4Ca.webp",42642,1,{"title":219,"description":220,"slug":221,"created_at":222,"publish_at":223,"formatted_created_at":224,"category":225,"links":226,"view_type":188,"video_url":87,"views":231,"likes":174,"lang":90,"comments_count":89,"is_pinned":111},"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":227,"image_webp":228,"thumb":229,"thumb_webp":230},"https://cdn.quasa.io/images/news/Wbx2NtS1CnTupgoQbpFMGspJ5jm4uob2hDOq33r0.jpg","https://cdn.quasa.io/images/news/Wbx2NtS1CnTupgoQbpFMGspJ5jm4uob2hDOq33r0.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/Wbx2NtS1CnTupgoQbpFMGspJ5jm4uob2hDOq33r0.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/Wbx2NtS1CnTupgoQbpFMGspJ5jm4uob2hDOq33r0.webp",41712,[233,234,235,236,237,238,239,240,241,242,243,244,245],{"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}]