[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"nav-categories":3,"article-ai2-unveils-open-coding-agents-sera-affordable-ai-for-real-world-codebases":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":97,"popular_news":166,"categories":237},{"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":86,"video_url":87,"views":88,"likes":89,"lang":90,"comments_count":89,"category":91},"Ai2 Unveils Open Coding Agents SERA: Affordable AI for Real-World Codebases","The Allen Institute for AI (Ai2) has launched SERA (Soft-verified Efficient Repository Agents), a family of open-source AI coding agents designed to tackle real-world codebases with unprecedented efficiency and adaptability.","SERA addresses key pain points in existing coding agents: high costs, lack of adaptability to private code, and the need for complex infrastructure.","These features make SERA accessible for small teams, requiring only standard supervised fine-tuning (SFT) without custom reinforcement learning (RL) tools","\u003Cp>The Allen Institute for AI (Ai2) has launched \u003Ca href=\"https://allenai.org/blog/open-coding-agents\">SERA (Soft-verified Efficient Repository Agents),\u003C/a> a family of open-source AI coding agents designed to tackle real-world codebases with unprecedented efficiency and adaptability.\u003C/p>\n\n\u003Cp>\u003Cpicture class=\"image-align-left\">\u003Csource srcset=\"https://cdn.quasa.io/photos/00/image-2026-02-07t141658593.webp\" type=\"image/webp\">\u003Cimg alt=\"Ai2 Unveils Open Coding Agents SERA: Affordable AI for Real-World Codebases\" class=\"image-align-left\" height=\"169\" src=\"https://cdn.quasa.io/photos/00/image-2026-02-07t141658593.jpg\" width=\"300\" />\u003C/picture>These models, ranging from 8B to 32B parameters (with a recent addition of SERA-14B), enable tasks like code generation, review, debugging, and explanation, all while being fine-tunable on private repositories.\u003C/p>\n\n\u003Cp>Built on the Qwen 3 backbone, SERA stands out for its low training costs &mdash; starting at just $400 &mdash; and its ability to outperform larger proprietary models in specialized scenarios.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Key Innovations in SERA\u003C/strong>\u003C/h4>\n\n\u003Cp>SERA addresses key pain points in existing coding agents: high costs, lack of adaptability to private code, and the need for complex infrastructure.\u003C/p>\n\n\u003Cp>\u003Cstrong>\u003Cpicture class=\"image-align-right\">\u003Csource srcset=\"https://cdn.quasa.io/photos/00/image-2026-02-07t141630100.webp\" type=\"image/webp\">\u003Cimg alt=\"Ai2 Unveils Open Coding Agents SERA: Affordable AI for Real-World Codebases\" class=\"image-align-right\" height=\"447\" src=\"https://cdn.quasa.io/photos/00/image-2026-02-07t141630100.jpg\" width=\"300\" />\u003C/picture>Its standout features include:\u003C/strong>\u003C/p>\n\n\u003Cul>\n\t\u003Cli>\u003Cstrong>Repository Specialization\u003C/strong>: Users can fine-tune SERA on private data, such as internal APIs or custom conventions, allowing smaller models to match or exceed larger &quot;teacher&quot; models like GLM-4.5-Air. For example, after training on 8,000 synthetic samples costing $1,300, SERA-32B surpassed GLM-4.5-Air on benchmarks for Django and SymPy repositories.\u003C/li>\n\t\u003Cli>\u003Cstrong>Soft-Verified Generation (SVG)\u003C/strong>: This method uses partially correct code patches for training, eliminating the need for full testing setups and scaling as effectively as &quot;hard-verified&quot; data.\u003C/li>\n\t\u003Cli>\u003Cstrong>Bug-Type Menu\u003C/strong>: Drawing from 51 common bug patterns, this generates diverse training prompts, creating thousands of agentic trajectories from repositories with limited functions.\u003C/li>\n\t\u003Cli>\u003Cstrong>High Workflow Fidelity\u003C/strong>: Prioritizes simulating real developer processes over perfect code accuracy, making it ideal for practical use.\u003C/li>\n\t\u003Cli>\u003Cstrong>Inference Speed\u003C/strong>: Optimized for high throughput, achieving up to 8,600 tokens per second on advanced NVIDIA hardware.\u003C/li>\n\u003C/ul>\n\n\u003Cp>These features make SERA accessible for small teams, requiring only standard supervised fine-tuning (SFT) without custom reinforcement learning (RL) tools.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Benchmark Performance\u003C/strong>\u003C/h4>\n\n\u003Cp>SERA excels on SWE-Bench Verified, a rigorous benchmark for coding agents. The flagship SERA-32B resolves 54.2% of tasks at 64K context, outperforming prior open-source models like Qwen3-Coder and matching proprietary ones such as Devstral Small 2 (50.0%) and GLM-4.5-Air (50.5%). The smaller SERA-8B achieves 29.4%, competitive in its weight class.\u003C/p>\n\n\u003Cp>\u003Cstrong>Here&#39;s a quick comparison on SWE-Bench Verified (resolve rates):\u003C/strong>\u003C/p>\n\n\u003Cp>\u003Cpicture>\u003Csource srcset=\"https://cdn.quasa.io/photos/00/2026-02-07_14-12-56.webp\" type=\"image/webp\">\u003Cimg alt=\"Ai2 Unveils Open Coding Agents SERA: Affordable AI for Real-World Codebases\" height=\"148\" src=\"https://cdn.quasa.io/photos/00/2026-02-07_14-12-56.png\" width=\"500\" />\u003C/picture>\u003C/p>\n\n\u003Cp>In repository-specific tests, SERA-32B hit 52.23% on Django and 51.11% on SymPy, edging out GLM-4.5-Air.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Training Efficiency: The Power of Synthetic Data\u003C/strong>\u003C/h4>\n\n\u003Cp>\u003Cpicture class=\"image-align-right\">\u003Csource srcset=\"https://cdn.quasa.io/photos/00/image-2026-02-07t141630870.webp\" type=\"image/webp\">\u003Cimg alt=\"Ai2 Unveils Open Coding Agents SERA: Affordable AI for Real-World Codebases\" class=\"image-align-right\" height=\"447\" src=\"https://cdn.quasa.io/photos/00/image-2026-02-07t141630870.jpg\" width=\"300\" />\u003C/picture>Ai2&#39;s breakthrough lies in cost-effective training using synthetic data generated via SVG and the bug-type menu. Reproducing top open-source performance costs around $400 &mdash; 57 times cheaper than SWE-smith and 26 times less than SkyRL.\u003C/p>\n\n\u003Cp>Scaling to rival industry leaders runs up to $12,000, still a fraction of traditional methods. Training occurs on modest hardware, like 2-16 NVIDIA GPUs, making it feasible for broader adoption.\u003C/p>\n\n\u003Cp>Also read:\u003C/p>\n\n\u003Cul>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/the-new-oil-is-human-cloudflare-and-wikimedia-pivot-to-ai-licensing\">The New Oil is Human: Cloudflare and Wikimedia Pivot to AI Licensing\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/nvidia-unveils-earth-2-the-world-s-first-fully-open-accelerated-ai-weather-climate-platform\">NVIDIA Unveils Earth-2: The World&#39;s First Fully Open, Accelerated AI Weather &amp; Climate Platform\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/tencent-s-hunyuanimage-3-0-instruct-the-thinking-multimodal-model-that-redefines-precise-image-editing\">Tencent&#39;s HunyuanImage 3.0-Instruct: The Thinking Multimodal Model That Redefines Precise Image Editing\u003C/a>\u003C/li>\n\t\u003Cli>\u003Ca href=\"https://quasa.io/media/good-news-for-spotify-apple-finally-approves-long-awaited-update\">Good News for Spotify: Apple Finally Approves Long-Awaited Update\u003C/a>\u003C/li>\n\u003C/ul>\n\n\u003Ch4>\u003Cstrong>Open-Source Release and Ease of Use\u003C/strong>\u003C/h4>\n\n\u003Cp>Everything is fully open-source on Hugging Face, including models, datasets (now in a model-agnostic format with verification metadata), training recipes, and inference code. Scripts for integration with Claude Code allow seamless setup &mdash; launch an inference server with just two lines of code. This democratizes AI coding agents, enabling anyone to create custom developers for their needs.\u003C/p>\n\n\u003Chr />\n\u003Ch4>\u003Cstrong>Looking Ahead\u003C/strong>\u003C/h4>\n\n\u003Cp>SERA represents a shift toward accessible, efficient AI for software engineering, lowering barriers for research and industry. By making high-performance agents reproducible on a budget, Ai2 fosters innovation in coding AI, potentially accelerating progress toward more advanced systems. As the community adopts these tools, we may see a surge in specialized agents tailored to diverse codebases worldwide.\u003C/p>","ai2-unveils-open-coding-agents-sera-affordable-ai-for-real-world-codebases","2026-02-07T13:21:24.000000Z","2026-02-17T03:11:00.000000Z","17.02.2026",{"image":82,"image_webp":83,"thumb":84,"thumb_webp":85},"https://cdn.quasa.io/images/news/3umo0wwcvD8e9OSdSRb6dDNesRJoAMoVM7b6rhx8.jpg","https://cdn.quasa.io/images/news/3umo0wwcvD8e9OSdSRb6dDNesRJoAMoVM7b6rhx8.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/3umo0wwcvD8e9OSdSRb6dDNesRJoAMoVM7b6rhx8.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/3umo0wwcvD8e9OSdSRb6dDNesRJoAMoVM7b6rhx8.webp","small",null,753,0,"en",{"id":42,"title":43,"slug":44,"meta_title":92,"meta_description":93,"meta_keywords":94,"deleted_at":87,"created_at":95,"updated_at":96,"lang":90},"Technology | AI Breakthroughs and Fresh News | QUASA","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","2026-04-22T15:05:32.000000Z",[98,113,126,139,152],{"title":99,"description":100,"slug":101,"created_at":102,"publish_at":103,"formatted_created_at":104,"category":105,"links":106,"view_type":86,"video_url":87,"views":111,"likes":89,"lang":90,"comments_count":89,"is_pinned":112},"Grok Voice Think Fast 1.0: The $3/Hour Voice AI That’s About to Replace Entire Call Centers","xAI just released Grok Voice Think Fast 1.0 — and the economics of voice AI have officially broken.","grok-voice-think-fast-1-0-the-3-hour-voice-ai-that-s-about-to-replace-entire-call-centers","2026-04-27T09:48:30.000000Z","2026-04-29T11:36:00.000000Z","29.04.2026",{"title":58,"slug":63},{"image":107,"image_webp":108,"thumb":109,"thumb_webp":110},"https://cdn.quasa.io/images/news/CaLws6TKSTHgTelvUiTRfEcaVyOIWzXu77AvSM8G.jpg","https://cdn.quasa.io/images/news/CaLws6TKSTHgTelvUiTRfEcaVyOIWzXu77AvSM8G.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/CaLws6TKSTHgTelvUiTRfEcaVyOIWzXu77AvSM8G.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/CaLws6TKSTHgTelvUiTRfEcaVyOIWzXu77AvSM8G.webp",23,false,{"title":114,"description":115,"slug":116,"created_at":117,"publish_at":118,"formatted_created_at":104,"category":119,"links":120,"view_type":86,"video_url":87,"views":125,"likes":89,"lang":90,"comments_count":89,"is_pinned":112},"DeepSeek-V4: The Open-Source Model That Just Made 1 Million Tokens Routine","DeepSeek has dropped the preview of DeepSeek-V4 — a direct, no-compromise open-source challenger to Claude, GPT-5, and Gemini.","deepseek-v4-the-open-source-model-that-just-made-1-million-tokens-routine","2026-04-27T09:31:03.000000Z","2026-04-29T09:20:00.000000Z",{"title":47,"slug":48},{"image":121,"image_webp":122,"thumb":123,"thumb_webp":124},"https://cdn.quasa.io/images/news/tbCZ87Xbns1do1OzsCcVqR7Hyb5TxUo0d57URt3Z.jpg","https://cdn.quasa.io/images/news/tbCZ87Xbns1do1OzsCcVqR7Hyb5TxUo0d57URt3Z.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/tbCZ87Xbns1do1OzsCcVqR7Hyb5TxUo0d57URt3Z.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/tbCZ87Xbns1do1OzsCcVqR7Hyb5TxUo0d57URt3Z.webp",37,{"title":127,"description":128,"slug":129,"created_at":130,"publish_at":131,"formatted_created_at":104,"category":132,"links":133,"view_type":86,"video_url":87,"views":138,"likes":89,"lang":90,"comments_count":89,"is_pinned":112},"OpenAI’s $10B DeployCo Play: Turning Private Equity Portfolios into an Enterprise AI Launchpad","OpenAI has just made its boldest move yet to crack the enterprise market. Instead of relying on traditional sales teams or pure software licensing, the company is betting big on implementation — and it’s doing so through a massive new joint venture called DeployCo.","openai-s-10b-deployco-play-turning-private-equity-portfolios-into-an-enterprise-ai-launchpad","2026-04-27T09:17:47.000000Z","2026-04-29T06:05:00.000000Z",{"title":43,"slug":44},{"image":134,"image_webp":135,"thumb":136,"thumb_webp":137},"https://cdn.quasa.io/images/news/FTRf04i8FQ4x4ILvUmSbpmzrD6LzgCZHmqpnMqUj.jpg","https://cdn.quasa.io/images/news/FTRf04i8FQ4x4ILvUmSbpmzrD6LzgCZHmqpnMqUj.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/FTRf04i8FQ4x4ILvUmSbpmzrD6LzgCZHmqpnMqUj.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/FTRf04i8FQ4x4ILvUmSbpmzrD6LzgCZHmqpnMqUj.webp",55,{"title":140,"description":141,"slug":142,"created_at":143,"publish_at":144,"formatted_created_at":104,"category":145,"links":146,"view_type":86,"video_url":87,"views":151,"likes":89,"lang":90,"comments_count":89,"is_pinned":112},"ComfyUI Cracks Open Real Human Faces in Seedance 2.0 — But Is It Really Fixed?","Anyone who has spent time wrestling with Seedance 2.0 in ComfyUI knows the pain. The model is incredibly powerful — cinematic camera control, native audio sync, buttery motion — but the built-in censorship made generating videos with actual human faces a nightmare.","comfyui-cracks-open-real-human-faces-in-seedance-2-0-but-is-it-really-fixed","2026-04-27T09:02:16.000000Z","2026-04-29T03:50:00.000000Z",{"title":27,"slug":28},{"image":147,"image_webp":148,"thumb":149,"thumb_webp":150},"https://cdn.quasa.io/images/news/isioWpVV24Sw2xRNu9cnOgRXYy28g1ASNCW254ZX.jpg","https://cdn.quasa.io/images/news/isioWpVV24Sw2xRNu9cnOgRXYy28g1ASNCW254ZX.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/isioWpVV24Sw2xRNu9cnOgRXYy28g1ASNCW254ZX.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/isioWpVV24Sw2xRNu9cnOgRXYy28g1ASNCW254ZX.webp",67,{"title":153,"description":154,"slug":155,"created_at":156,"publish_at":157,"formatted_created_at":158,"category":159,"links":160,"view_type":86,"video_url":87,"views":165,"likes":89,"lang":90,"comments_count":89,"is_pinned":112},"Sierra AI Just Reinvented the Software Engineering Interview — And It’s Brilliant","Sierra AI — the enterprise AI startup co-founded by Bret Taylor (OpenAI board chairman and former Co-CEO of Salesforce) — just published a quietly revolutionary blog post titled The AI-native interview.","sierra-ai-just-reinvented-the-software-engineering-interview-and-it-s-brilliant","2026-04-23T17:49:41.000000Z","2026-04-28T11:39:00.000000Z","28.04.2026",{"title":19,"slug":20},{"image":161,"image_webp":162,"thumb":163,"thumb_webp":164},"https://cdn.quasa.io/images/news/DJ894jC3gLB6PUzLn5skZgOEm7fBiFooBdGQUwpC.jpg","https://cdn.quasa.io/images/news/DJ894jC3gLB6PUzLn5skZgOEm7fBiFooBdGQUwpC.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/DJ894jC3gLB6PUzLn5skZgOEm7fBiFooBdGQUwpC.jpg","https://cdn.quasa.io/thumbs/news-thumb/images/news/DJ894jC3gLB6PUzLn5skZgOEm7fBiFooBdGQUwpC.webp",155,[167,180,196,208,223],{"title":168,"description":169,"slug":170,"created_at":171,"publish_at":172,"formatted_created_at":173,"category":174,"links":175,"view_type":86,"video_url":87,"views":178,"likes":179,"lang":90,"comments_count":89,"is_pinned":112},"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":176,"image_webp":87,"thumb":177,"thumb_webp":177},"https://cdn.quasa.io/images/news/mVsXPTMuHZuI7UXCsENgL1Qwp1uSOf7Rz3uVPMfm.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/mVsXPTMuHZuI7UXCsENgL1Qwp1uSOf7Rz3uVPMfm.webp",72109,2,{"title":181,"description":182,"slug":183,"created_at":184,"publish_at":185,"formatted_created_at":186,"category":187,"links":188,"view_type":193,"video_url":87,"views":194,"likes":195,"lang":90,"comments_count":89,"is_pinned":112},"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":189,"image_webp":190,"thumb":191,"thumb_webp":192},"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",71890,4,{"title":197,"description":198,"slug":199,"created_at":200,"publish_at":201,"formatted_created_at":202,"category":203,"links":204,"view_type":86,"video_url":87,"views":207,"likes":195,"lang":90,"comments_count":89,"is_pinned":112},"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":205,"image_webp":87,"thumb":206,"thumb_webp":206},"https://cdn.quasa.io/images/news/EOsQhSW3VXyG7a6NPdE1oZd00xfJXe3bjY5aJGb7.webp","https://cdn.quasa.io/thumbs/news-thumb/images/news/EOsQhSW3VXyG7a6NPdE1oZd00xfJXe3bjY5aJGb7.webp",69474,{"title":209,"description":210,"slug":211,"created_at":212,"publish_at":213,"formatted_created_at":214,"category":215,"links":216,"view_type":86,"video_url":87,"views":221,"likes":179,"lang":90,"comments_count":222,"is_pinned":112},"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":217,"image_webp":218,"thumb":219,"thumb_webp":220},"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",43368,1,{"title":224,"description":225,"slug":226,"created_at":227,"publish_at":228,"formatted_created_at":229,"category":230,"links":231,"view_type":193,"video_url":87,"views":236,"likes":179,"lang":90,"comments_count":89,"is_pinned":112},"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":232,"image_webp":233,"thumb":234,"thumb_webp":235},"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",42381,[238,239,240,241,242,243,244,245,246,247,248,249,250],{"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}]