{"id":71995,"date":"2026-02-18T11:50:07","date_gmt":"2026-02-18T11:50:07","guid":{"rendered":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/"},"modified":"2026-02-18T11:50:07","modified_gmt":"2026-02-18T11:50:07","slug":"cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone","status":"publish","type":"post","link":"https:\/\/youzum.net\/ja\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/","title":{"rendered":"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone"},"content":{"rendered":"<p>Cohere AI Labs has released <strong>Tiny Aya<\/strong>, a family of small language models (SLMs) that redefines multilingual performance. While many models scale by increasing parameters, Tiny Aya uses a <strong>3.35B-parameter<\/strong> architecture to deliver state-of-the-art translation and generation across <strong>70 languages<\/strong>.<\/p>\n<p>The release includes 5 models: <strong>Tiny Aya Base<\/strong> (pretrained), <strong>Tiny Aya Global<\/strong> (balanced instruction-tuned), and three region-specific variants\u2014<strong>Earth<\/strong> (Africa\/West Asia), <strong>Fire<\/strong> (South Asia), and <strong>Water<\/strong> (Asia-Pacific\/Europe).<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full is-resized\"><img fetchpriority=\"high\" decoding=\"async\" width=\"1000\" height=\"563\" data-attachment-id=\"77954\" data-permalink=\"https:\/\/www.marktechpost.com\/2026\/02\/17\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/image-322\/\" data-orig-file=\"https:\/\/www.marktechpost.com\/wp-content\/uploads\/2026\/02\/image-12.png\" data-orig-size=\"1000,563\" data-comments-opened=\"1\" data-image-meta='{\"aperture\":\"0\",\"credit\":\"\",\"camera\":\"\",\"caption\":\"\",\"created_timestamp\":\"0\",\"copyright\":\"\",\"focal_length\":\"0\",\"iso\":\"0\",\"shutter_speed\":\"0\",\"title\":\"\",\"orientation\":\"0\"}' data-image-title=\"image\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/www.marktechpost.com\/wp-content\/uploads\/2026\/02\/image-12-300x169.png\" data-large-file=\"https:\/\/www.marktechpost.com\/wp-content\/uploads\/2026\/02\/image-12.png\" src=\"https:\/\/www.marktechpost.com\/wp-content\/uploads\/2026\/02\/image-12.png\" alt=\"\" class=\"wp-image-77954\" \/><figcaption class=\"wp-element-caption\">https:\/\/cohere.com\/blog\/cohere-labs-tiny-aya<\/figcaption><\/figure>\n<\/div>\n<h3 class=\"wp-block-heading\"><strong>The Architecture<\/strong><\/h3>\n<p>Tiny Aya is built on a dense decoder-only Transformer architecture. <strong>Key specs include:<\/strong><\/p>\n<ul class=\"wp-block-list\">\n<li><strong>Parameters<\/strong>: 3.35B total (2.8B non-embedding)<\/li>\n<li><strong>Layers<\/strong>: 36<\/li>\n<li><strong>Vocabulary<\/strong>: 262k tokenizer designed for equitable language representation.<\/li>\n<li><strong>Attention<\/strong>: Interleaved sliding window and full attention (3:1 ratio) with Grouped Query Attention (GQA).<\/li>\n<li><strong>Context<\/strong>: 8192 tokens for input and output.<\/li>\n<\/ul>\n<p>The model was pretrained on <strong>6T tokens<\/strong> using a Warmup-Stable-Decay (WSD) schedule. To maintain stability, the team used <strong>SwiGLU<\/strong> activations and removed all biases from dense layers.<\/p>\n<h3 class=\"wp-block-heading\"><strong>Advanced Post-training: FUSION and SimMerge<\/strong><\/h3>\n<p>To bridge the gap in low-resource languages, Cohere used a synthetic data pipeline<sup><\/sup>.<\/p>\n<ol start=\"1\" class=\"wp-block-list\">\n<li><strong>Fusion-of-N (FUSION)<\/strong>: Prompts are sent to a \u2018team of teachers\u2019 (COMMAND A, GEMMA3-27B-IT, DEEPSEEK-V3). A judge LLM, the <strong>Fusor<\/strong>, extracts and aggregates the strongest components of their responses.<\/li>\n<li><strong>Region Specialization<\/strong>: Models were finetuned on 5 regional clusters (e.g., South Asia, Africa).<\/li>\n<li><strong>SimMerge<\/strong>: To prevent \u2018catastrophic forgetting\u2019 of global safety, regional checkpoints were merged with the global model using <strong>SimMerge<\/strong>, which selects the best merge operators based on similarity signals.<\/li>\n<\/ol>\n<h3 class=\"wp-block-heading\"><strong>Performance Benchmarks<\/strong><\/h3>\n<p><strong>Tiny Aya Global consistently beats larger or same-scale competitors in multilingual tasks:<\/strong><\/p>\n<ul class=\"wp-block-list\">\n<li><strong>Translation<\/strong>: It outperforms <strong>GEMMA3-4B<\/strong> in <strong>46 of 61 languages<\/strong> on <strong>WMT24++<\/strong>.<\/li>\n<li><strong>Reasoning<\/strong>: In the <strong>GlobalMGSM<\/strong> (math) benchmark for African languages, Tiny Aya achieved <strong>39.2% accuracy<\/strong>, dwarfing <strong>GEMMA3-4B<\/strong> (17.6%) and <strong>QWEN3-4B<\/strong> (6.25%).<\/li>\n<li><strong>Safety<\/strong>: It holds the highest mean safe response rate (<strong>91.1%<\/strong>) on <strong>MultiJail<\/strong>.<\/li>\n<li><strong>Language Integrity<\/strong>: The model achieves <strong>94% language accuracy<\/strong>, meaning it rarely switches to English when asked to reply in another language.<\/li>\n<\/ul>\n<h3 class=\"wp-block-heading\"><strong>On-Device Deployment<\/strong><\/h3>\n<p>Tiny Aya is optimized for edge computing. Using <strong>4-bit quantization (Q4_K_M)<\/strong>, the model fits in a <strong>2.14 GB<\/strong> memory footprint.<\/p>\n<ul class=\"wp-block-list\">\n<li><strong>iPhone 13<\/strong>: 10 tokens\/s.<\/li>\n<li><strong>iPhone 17 Pro<\/strong>: 32 tokens\/s.<\/li>\n<\/ul>\n<p>This quantization scheme results in a minimal <strong>1.4-point<\/strong> drop in generation quality, making it a viable solution for offline, private, and localized AI applications.<\/p>\n<h3 class=\"wp-block-heading\"><strong>Key Takeaways<\/strong><\/h3>\n<ul class=\"wp-block-list\">\n<li><strong>Efficient Multilingual Power<\/strong>: Tiny Aya is a <strong>3.35B-parameter<\/strong> model family that delivers state-of-the-art translation and high-quality generation across <strong>70 languages<\/strong>. It proves that massive scale is not required for strong multilingual performance if models are designed with intentional data curation.<\/li>\n<li><strong>Innovative Training Pipeline<\/strong>: The models were developed using a novel strategy involving <strong>Fusion-of-N (FUSION)<\/strong>, where a \u2018team of teachers\u2019 (like Command A and DeepSeek-V3) generated synthetic data. A judge model then aggregated the strongest components to ensure high-quality training signals even for low-resource languages.<\/li>\n<li><strong>Regional Specialization via Merging<\/strong>: Cohere released specialized variants\u2014<strong>Tiny Aya Earth, Fire, and Water<\/strong>\u2014which are tuned for specific regions like Africa, South Asia, and the Asia-Pacific. These were created by merging regional fine-tuned models with a global model using <strong>SimMerge<\/strong> to preserve safety while boosting local language performance.<\/li>\n<li><strong>Superior Benchmark Performance<\/strong>: Tiny Aya Global outperforms competitors like <strong>Gemma3-4B<\/strong> in translation quality for <strong>46 of 61 languages<\/strong> on WMT24++. It also significantly reduces disparities in mathematical reasoning for African languages, achieving <strong>39.2% accuracy<\/strong> compared to Gemma3-4B\u2019s 17.6%.<\/li>\n<li><strong>Optimized for On-Device Deployment<\/strong>: The model is highly portable and runs efficiently on edge devices; it achieves <strong>~10 tokens\/s<\/strong> on an iPhone 13 and <strong>32 tokens\/s<\/strong> on an iPhone 17 Pro using <strong>Q4_K_M quantization<\/strong>. This 4-bit quantization format maintains high quality with only a minimal <strong>1.4-point<\/strong> degradation.<\/li>\n<\/ul>\n<hr class=\"wp-block-separator has-alpha-channel-opacity\" \/>\n<p>Check out the\u00a0<strong><a href=\"https:\/\/cohere.com\/blog\/cohere-labs-tiny-aya\" target=\"_blank\" rel=\"noreferrer noopener\">Technical details<\/a>, <a href=\"https:\/\/github.com\/Cohere-Labs\/tiny-aya-tech-report\/blob\/main\/tiny_aya_tech_report.pdf\" target=\"_blank\" rel=\"noreferrer noopener\">Paper<\/a>, <a href=\"https:\/\/huggingface.co\/collections\/CohereLabs\/tiny-aya?ref=cohere.com\/blog\" target=\"_blank\" rel=\"noreferrer noopener\">Model Weights<\/a> and <a href=\"https:\/\/huggingface.co\/spaces\/CohereLabs\/tiny-aya?ref=cohere.com%2Fblog\" target=\"_blank\" rel=\"noreferrer noopener\">Playground<\/a>.\u00a0<\/strong>Also,\u00a0feel free to follow us on\u00a0<strong><a href=\"https:\/\/x.com\/intent\/follow?screen_name=marktechpost\" target=\"_blank\" rel=\"noreferrer noopener\"><mark>Twitter<\/mark><\/a><\/strong>\u00a0and don\u2019t forget to join our\u00a0<strong><a href=\"https:\/\/www.reddit.com\/r\/machinelearningnews\/\" target=\"_blank\" rel=\"noreferrer noopener\">100k+ ML SubReddit<\/a><\/strong>\u00a0and Subscribe to\u00a0<strong><a href=\"https:\/\/www.aidevsignals.com\/\" target=\"_blank\" rel=\"noreferrer noopener\">our Newsletter<\/a><\/strong>. Wait! are you on telegram?\u00a0<strong><a href=\"https:\/\/t.me\/machinelearningresearchnews\" target=\"_blank\" rel=\"noreferrer noopener\">now you can join us on telegram as well.<\/a><\/strong><\/p>\n<p>The post <a href=\"https:\/\/www.marktechpost.com\/2026\/02\/17\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\">Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone<\/a> appeared first on <a href=\"https:\/\/www.marktechpost.com\/\">MarkTechPost<\/a>.<\/p>","protected":false},"excerpt":{"rendered":"<p>Cohere AI Labs has released Tiny Aya, a family of small language models (SLMs) that redefines multilingual performance. While many models scale by increasing parameters, Tiny Aya uses a 3.35B-parameter architecture to deliver state-of-the-art translation and generation across 70 languages. The release includes 5 models: Tiny Aya Base (pretrained), Tiny Aya Global (balanced instruction-tuned), and three region-specific variants\u2014Earth (Africa\/West Asia), Fire (South Asia), and Water (Asia-Pacific\/Europe). https:\/\/cohere.com\/blog\/cohere-labs-tiny-aya The Architecture Tiny Aya is built on a dense decoder-only Transformer architecture. Key specs include: Parameters: 3.35B total (2.8B non-embedding) Layers: 36 Vocabulary: 262k tokenizer designed for equitable language representation. Attention: Interleaved sliding window and full attention (3:1 ratio) with Grouped Query Attention (GQA). Context: 8192 tokens for input and output. The model was pretrained on 6T tokens using a Warmup-Stable-Decay (WSD) schedule. To maintain stability, the team used SwiGLU activations and removed all biases from dense layers. Advanced Post-training: FUSION and SimMerge To bridge the gap in low-resource languages, Cohere used a synthetic data pipeline. Fusion-of-N (FUSION): Prompts are sent to a \u2018team of teachers\u2019 (COMMAND A, GEMMA3-27B-IT, DEEPSEEK-V3). A judge LLM, the Fusor, extracts and aggregates the strongest components of their responses. Region Specialization: Models were finetuned on 5 regional clusters (e.g., South Asia, Africa). SimMerge: To prevent \u2018catastrophic forgetting\u2019 of global safety, regional checkpoints were merged with the global model using SimMerge, which selects the best merge operators based on similarity signals. Performance Benchmarks Tiny Aya Global consistently beats larger or same-scale competitors in multilingual tasks: Translation: It outperforms GEMMA3-4B in 46 of 61 languages on WMT24++. Reasoning: In the GlobalMGSM (math) benchmark for African languages, Tiny Aya achieved 39.2% accuracy, dwarfing GEMMA3-4B (17.6%) and QWEN3-4B (6.25%). Safety: It holds the highest mean safe response rate (91.1%) on MultiJail. Language Integrity: The model achieves 94% language accuracy, meaning it rarely switches to English when asked to reply in another language. On-Device Deployment Tiny Aya is optimized for edge computing. Using 4-bit quantization (Q4_K_M), the model fits in a 2.14 GB memory footprint. iPhone 13: 10 tokens\/s. iPhone 17 Pro: 32 tokens\/s. This quantization scheme results in a minimal 1.4-point drop in generation quality, making it a viable solution for offline, private, and localized AI applications. Key Takeaways Efficient Multilingual Power: Tiny Aya is a 3.35B-parameter model family that delivers state-of-the-art translation and high-quality generation across 70 languages. It proves that massive scale is not required for strong multilingual performance if models are designed with intentional data curation. Innovative Training Pipeline: The models were developed using a novel strategy involving Fusion-of-N (FUSION), where a \u2018team of teachers\u2019 (like Command A and DeepSeek-V3) generated synthetic data. A judge model then aggregated the strongest components to ensure high-quality training signals even for low-resource languages. Regional Specialization via Merging: Cohere released specialized variants\u2014Tiny Aya Earth, Fire, and Water\u2014which are tuned for specific regions like Africa, South Asia, and the Asia-Pacific. These were created by merging regional fine-tuned models with a global model using SimMerge to preserve safety while boosting local language performance. Superior Benchmark Performance: Tiny Aya Global outperforms competitors like Gemma3-4B in translation quality for 46 of 61 languages on WMT24++. It also significantly reduces disparities in mathematical reasoning for African languages, achieving 39.2% accuracy compared to Gemma3-4B\u2019s 17.6%. Optimized for On-Device Deployment: The model is highly portable and runs efficiently on edge devices; it achieves ~10 tokens\/s on an iPhone 13 and 32 tokens\/s on an iPhone 17 Pro using Q4_K_M quantization. This 4-bit quantization format maintains high quality with only a minimal 1.4-point degradation. Check out the\u00a0Technical details, Paper, Model Weights and Playground.\u00a0Also,\u00a0feel free to follow us on\u00a0Twitter\u00a0and don\u2019t forget to join our\u00a0100k+ ML SubReddit\u00a0and Subscribe to\u00a0our Newsletter. Wait! are you on telegram?\u00a0now you can join us on telegram as well. The post Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone appeared first on MarkTechPost.<\/p>","protected":false},"author":2,"featured_media":71996,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"pmpro_default_level":"","site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"_pvb_checkbox_block_on_post":false,"footnotes":""},"categories":[52,5,7,1],"tags":[],"class_list":["post-71995","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-club","category-committee","category-news","category-uncategorized","pmpro-has-access"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v25.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone - YouZum<\/title>\n<meta name=\"description\" content=\"\u0e01\u0e34\u0e08\u0e01\u0e23\u0e23\u0e21\u0e40\u0e01\u0e35\u0e48\u0e22\u0e27\u0e01\u0e31\u0e1a\u0e42\u0e14\u0e23\u0e19\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/youzum.net\/ja\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\" \/>\n<meta property=\"og:locale\" content=\"ja_JP\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone - YouZum\" \/>\n<meta property=\"og:description\" content=\"\u0e01\u0e34\u0e08\u0e01\u0e23\u0e23\u0e21\u0e40\u0e01\u0e35\u0e48\u0e22\u0e27\u0e01\u0e31\u0e1a\u0e42\u0e14\u0e23\u0e19\" \/>\n<meta property=\"og:url\" content=\"https:\/\/youzum.net\/ja\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\" \/>\n<meta property=\"og:site_name\" content=\"YouZum\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/DroneAssociationTH\/\" \/>\n<meta property=\"article:published_time\" content=\"2026-02-18T11:50:07+00:00\" \/>\n<meta name=\"author\" content=\"admin NU\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u57f7\u7b46\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin NU\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u63a8\u5b9a\u8aad\u307f\u53d6\u308a\u6642\u9593\" \/>\n\t<meta name=\"twitter:data2\" content=\"3\u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\"},\"author\":{\"name\":\"admin NU\",\"@id\":\"https:\/\/yousum.gpucore.co\/#\/schema\/person\/97fa48242daf3908e4d9a5f26f4a059c\"},\"headline\":\"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone\",\"datePublished\":\"2026-02-18T11:50:07+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\"},\"wordCount\":678,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/yousum.gpucore.co\/#organization\"},\"image\":{\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp\",\"articleSection\":[\"AI\",\"Committee\",\"News\",\"Uncategorized\"],\"inLanguage\":\"ja\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\",\"url\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\",\"name\":\"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone - YouZum\",\"isPartOf\":{\"@id\":\"https:\/\/yousum.gpucore.co\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp\",\"datePublished\":\"2026-02-18T11:50:07+00:00\",\"description\":\"\u0e01\u0e34\u0e08\u0e01\u0e23\u0e23\u0e21\u0e40\u0e01\u0e35\u0e48\u0e22\u0e27\u0e01\u0e31\u0e1a\u0e42\u0e14\u0e23\u0e19\",\"breadcrumb\":{\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#breadcrumb\"},\"inLanguage\":\"ja\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"ja\",\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage\",\"url\":\"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp\",\"contentUrl\":\"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp\",\"width\":1000,\"height\":563},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/youzum.net\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/yousum.gpucore.co\/#website\",\"url\":\"https:\/\/yousum.gpucore.co\/\",\"name\":\"YouSum\",\"description\":\"\",\"publisher\":{\"@id\":\"https:\/\/yousum.gpucore.co\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/yousum.gpucore.co\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"ja\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/yousum.gpucore.co\/#organization\",\"name\":\"Drone Association Thailand\",\"url\":\"https:\/\/yousum.gpucore.co\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ja\",\"@id\":\"https:\/\/yousum.gpucore.co\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/youzum.net\/wp-content\/uploads\/2024\/11\/tranparent-logo.png\",\"contentUrl\":\"https:\/\/youzum.net\/wp-content\/uploads\/2024\/11\/tranparent-logo.png\",\"width\":300,\"height\":300,\"caption\":\"Drone Association Thailand\"},\"image\":{\"@id\":\"https:\/\/yousum.gpucore.co\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/www.facebook.com\/DroneAssociationTH\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/yousum.gpucore.co\/#\/schema\/person\/97fa48242daf3908e4d9a5f26f4a059c\",\"name\":\"admin NU\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ja\",\"@id\":\"https:\/\/yousum.gpucore.co\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/youzum.net\/wp-content\/uploads\/avatars\/2\/1746849356-bpfull.png\",\"contentUrl\":\"https:\/\/youzum.net\/wp-content\/uploads\/avatars\/2\/1746849356-bpfull.png\",\"caption\":\"admin NU\"},\"url\":\"https:\/\/youzum.net\/ja\/members\/adminnu\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone - YouZum","description":"\u0e01\u0e34\u0e08\u0e01\u0e23\u0e23\u0e21\u0e40\u0e01\u0e35\u0e48\u0e22\u0e27\u0e01\u0e31\u0e1a\u0e42\u0e14\u0e23\u0e19","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/youzum.net\/ja\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/","og_locale":"ja_JP","og_type":"article","og_title":"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone - YouZum","og_description":"\u0e01\u0e34\u0e08\u0e01\u0e23\u0e23\u0e21\u0e40\u0e01\u0e35\u0e48\u0e22\u0e27\u0e01\u0e31\u0e1a\u0e42\u0e14\u0e23\u0e19","og_url":"https:\/\/youzum.net\/ja\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/","og_site_name":"YouZum","article_publisher":"https:\/\/www.facebook.com\/DroneAssociationTH\/","article_published_time":"2026-02-18T11:50:07+00:00","author":"admin NU","twitter_card":"summary_large_image","twitter_misc":{"\u57f7\u7b46\u8005":"admin NU","\u63a8\u5b9a\u8aad\u307f\u53d6\u308a\u6642\u9593":"3\u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#article","isPartOf":{"@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/"},"author":{"name":"admin NU","@id":"https:\/\/yousum.gpucore.co\/#\/schema\/person\/97fa48242daf3908e4d9a5f26f4a059c"},"headline":"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone","datePublished":"2026-02-18T11:50:07+00:00","mainEntityOfPage":{"@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/"},"wordCount":678,"commentCount":0,"publisher":{"@id":"https:\/\/yousum.gpucore.co\/#organization"},"image":{"@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage"},"thumbnailUrl":"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp","articleSection":["AI","Committee","News","Uncategorized"],"inLanguage":"ja","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/","url":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/","name":"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone - YouZum","isPartOf":{"@id":"https:\/\/yousum.gpucore.co\/#website"},"primaryImageOfPage":{"@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage"},"image":{"@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage"},"thumbnailUrl":"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp","datePublished":"2026-02-18T11:50:07+00:00","description":"\u0e01\u0e34\u0e08\u0e01\u0e23\u0e23\u0e21\u0e40\u0e01\u0e35\u0e48\u0e22\u0e27\u0e01\u0e31\u0e1a\u0e42\u0e14\u0e23\u0e19","breadcrumb":{"@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#breadcrumb"},"inLanguage":"ja","potentialAction":[{"@type":"ReadAction","target":["https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/"]}]},{"@type":"ImageObject","inLanguage":"ja","@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#primaryimage","url":"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp","contentUrl":"https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp","width":1000,"height":563},{"@type":"BreadcrumbList","@id":"https:\/\/youzum.net\/cohere-releases-tiny-aya-a-3b-parameter-small-language-model-that-supports-70-languages-and-runs-locally-even-on-a-phone\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/youzum.net\/"},{"@type":"ListItem","position":2,"name":"Cohere Releases Tiny Aya: A 3B-Parameter Small Language Model that Supports 70 Languages and Runs Locally Even on a Phone"}]},{"@type":"WebSite","@id":"https:\/\/yousum.gpucore.co\/#website","url":"https:\/\/yousum.gpucore.co\/","name":"YouSum","description":"","publisher":{"@id":"https:\/\/yousum.gpucore.co\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/yousum.gpucore.co\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"ja"},{"@type":"Organization","@id":"https:\/\/yousum.gpucore.co\/#organization","name":"Drone Association Thailand","url":"https:\/\/yousum.gpucore.co\/","logo":{"@type":"ImageObject","inLanguage":"ja","@id":"https:\/\/yousum.gpucore.co\/#\/schema\/logo\/image\/","url":"https:\/\/youzum.net\/wp-content\/uploads\/2024\/11\/tranparent-logo.png","contentUrl":"https:\/\/youzum.net\/wp-content\/uploads\/2024\/11\/tranparent-logo.png","width":300,"height":300,"caption":"Drone Association Thailand"},"image":{"@id":"https:\/\/yousum.gpucore.co\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/DroneAssociationTH\/"]},{"@type":"Person","@id":"https:\/\/yousum.gpucore.co\/#\/schema\/person\/97fa48242daf3908e4d9a5f26f4a059c","name":"admin NU","image":{"@type":"ImageObject","inLanguage":"ja","@id":"https:\/\/yousum.gpucore.co\/#\/schema\/person\/image\/","url":"https:\/\/youzum.net\/wp-content\/uploads\/avatars\/2\/1746849356-bpfull.png","contentUrl":"https:\/\/youzum.net\/wp-content\/uploads\/avatars\/2\/1746849356-bpfull.png","caption":"admin NU"},"url":"https:\/\/youzum.net\/ja\/members\/adminnu\/"}]}},"rttpg_featured_image_url":{"full":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp",1000,563,false],"landscape":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp",1000,563,false],"portraits":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp",1000,563,false],"thumbnail":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ-150x150.webp",150,150,true],"medium":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ-300x169.webp",300,169,true],"large":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp",1000,563,false],"1536x1536":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp",1000,563,false],"2048x2048":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ.webp",1000,563,false],"trp-custom-language-flag":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ-18x10.webp",18,10,true],"woocommerce_thumbnail":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ-300x300.webp",300,300,true],"woocommerce_single":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ-600x338.webp",600,338,true],"woocommerce_gallery_thumbnail":["https:\/\/youzum.net\/wp-content\/uploads\/2026\/02\/image-12-G9JOcJ-100x100.webp",100,100,true]},"rttpg_author":{"display_name":"admin NU","author_link":"https:\/\/youzum.net\/ja\/members\/adminnu\/"},"rttpg_comment":0,"rttpg_category":"<a href=\"https:\/\/youzum.net\/ja\/category\/ai-club\/\" rel=\"category tag\">AI<\/a> <a href=\"https:\/\/youzum.net\/ja\/category\/committee\/\" rel=\"category tag\">Committee<\/a> <a href=\"https:\/\/youzum.net\/ja\/category\/news\/\" rel=\"category tag\">News<\/a> <a href=\"https:\/\/youzum.net\/ja\/category\/uncategorized\/\" rel=\"category tag\">Uncategorized<\/a>","rttpg_excerpt":"Cohere AI Labs has released Tiny Aya, a family of small language models (SLMs) that redefines multilingual performance. While many models scale by increasing parameters, Tiny Aya uses a 3.35B-parameter architecture to deliver state-of-the-art translation and generation across 70 languages. The release includes 5 models: Tiny Aya Base (pretrained), Tiny Aya Global (balanced instruction-tuned), and&hellip;","_links":{"self":[{"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/posts\/71995","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/comments?post=71995"}],"version-history":[{"count":0,"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/posts\/71995\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/media\/71996"}],"wp:attachment":[{"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/media?parent=71995"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/categories?post=71995"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/youzum.net\/ja\/wp-json\/wp\/v2\/tags?post=71995"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}