{"id":2245057,"date":"2026-11-06T13:02:16","date_gmt":"2026-11-06T10:02:16","guid":{"rendered":"https:\/\/analyse.optim.biz\/?p=2245057"},"modified":"2026-11-06T13:02:16","modified_gmt":"2026-11-06T10:02:16","slug":"underdog-shows-off-a-32mb-speech-model-in-new-demo","status":"publish","type":"post","link":"https:\/\/analyse.optim.biz\/?p=2245057","title":{"rendered":"Underdog shows off a 32MB speech model in new demo"},"content":{"rendered":"<p>[analyse_image type=&#8221;featured&#8221; src=&#8221;https:\/\/techcrunch.com\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-05-at-7.02.34-PM.jpg?resize=1200,542&#8243;]<\/p>\n<div class=\"entry-content wp-block-post-content is-layout-constrained wp-block-post-content-is-layout-constrained\">\n<p id=\"speakable-summary\" class=\"wp-block-paragraph\">AI companies across the world are racing to make bigger and better models, but the problem is all of them need massive amounts of processing power. On the other end, some researchers and startups are creating smaller models that can run on edge devices, like your phone or computer.<\/p>\n<p class=\"wp-block-paragraph\">One such startup, Underdog, has released a game built with its new voice model called Woof that\u2019s only 32MB large.<\/p>\n<div class=\"tc-event-inline-cta\">\n<div class=\"hellobar_disrupt_promo_slot\"><\/div>\n<\/div>\n<p class=\"wp-block-paragraph\">The game features a room with a retro tech theme, a typewriter, and a CRT monitor. There is a desk lamp, which makes a sound when you click its switch. The intro warns users about \u201cintrusive\u201d cloud providers in the town \u2014 ostensibly a jab at cloud-based models.<\/p>\n<p class=\"wp-block-paragraph\">You can interact with the typewriter and music tuner.<\/p>\n<figure class=\"wp-block-embed is-type-rich is-provider-x wp-block-embed-x\">\n<div class=\"wp-block-embed__wrapper\">\n<blockquote class=\"twitter-tweet\" data-width=\"500\" data-dnt=\"true\">\n<p lang=\"en\" dir=\"ltr\">We made a game for our new frontier text-to-speech model, Woof. 31.7 MB, so small that it runs in your browser  <\/p>\n<p>Play the game on X\u2026 all voices are Woof. Defeat Darkcloud for an invite https:\/\/t.co\/ImUFYDCBQG pic.twitter.com\/tthHx7ZGBK<\/p>\n<p>\u2014 Underdog AI (@UnderdogAI) November 5, 2026<\/p><\/blockquote>\n<\/div>\n<\/figure>\n<p class=\"wp-block-paragraph\">You can type anything on the typewriter using your keyboard, and the game will speak it back. You can also rotate the tuner to generate different kinds of music.<\/p>\n<p class=\"wp-block-paragraph\">All of this works within the browser through Underdog\u2019s speech generation model, which, the company says, can run on most devices without any internet connection. The model has 31 million parameters and is open for everyone to download. <\/p>\n<p class=\"wp-block-paragraph\">While such a small model likely won\u2019t replace bigger models, it may still be enough to drive local assistant experiences or create personalized podcasts.<\/p>\n<div class=\"jw-player-inline-promo wp-block-techcrunch-jw-player-embed\">\n<div id=\"jwplayer_6af0e26242f27\"><\/div>\n<\/div>\n<p class=\"wp-block-paragraph\">This demo comes weeks after Underdog launched Bark, a 167MB speech-to-text model also designed to run on edge devices for transcription and dictation.<\/p>\n<p class=\"wp-block-paragraph\">Sigil Wen, the founder of Underdog, told TechCrunch that he started building the app because, instead of using notetakers like Granola, he wanted to build his own system with AI models that can run on users\u2019 devices and let them own their data.<\/p>\n<p class=\"wp-block-paragraph\">\u201cI think we will be able to run most capable models of today on the devices you already own within months. We just want to train the best open models that are American-made, that are open, that are capable and allow people to have private AI,\u201d he said, referring to his manifesto.<\/p>\n<div class=\"ad-unit ad-unit--mobile ad-unit--has-placeholder ad-unit--display wp-block-tc-ads-ad-slot has-background has-gray-100-background-color\">\n<div class=\"ad-unit__ad\" id=\"us-tc-ros-mw-mid-center\" data-unitcode=\"us_tc_ros_mw_mid_center\" data-test=\"us_tc_ros_mw_mid_center\">\n\t<\/div>\n<\/div>\n<div class=\"ad-unit ad-unit--desktop ad-unit--native wp-block-tc-ads-ad-slot\">\n<div class=\"ad-unit__ad\" id=\"us-tc-ros-dt-native-midarticle\" data-unitcode=\"us_tc_ros_dt_native_midarticle\" data-test=\"us_tc_ros_dt_native_midarticle\">\n\t<\/div>\n<\/div>\n<p class=\"wp-block-paragraph\">In addition to the two voice models mentioned above, Underdog has also released a 27-billion-parameter model that can run on laptops, along with a waitlisted app that the company is positioning as a locally hosted alternative to Muse and Instinct.<\/p>\n<\/div>\n<p>[analyse_source url=&#8221;https:\/\/techcrunch.com\/2026\/11\/06\/underdog-shows-off-a-32mb-speech-model-in-new-demo\/&#8221;]<\/p>\n","protected":false},"excerpt":{"rendered":"<p>[analyse_image type=&#8221;featured&#8221; src=&#8221;https:\/\/techcrunch.com\/wp-content\/uploads\/2026\/11\/Screenshot-2026-11-05-at-7.02.34-PM.jpg?resize=1200,542&#8243;] AI companies across the world are racing to make bigger and better models, but the problem is all of them need massive amounts of processing power. On the other end, some researchers and startups are creating smaller models that can run on edge devices, like your phone or computer. One such startup, [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[226,62],"class_list":["post-2245057","post","type-post","status-publish","format-standard","hentry","category-politics","tag-crawlmanager","tag-techcrunch-com"],"_links":{"self":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts\/2245057","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2245057"}],"version-history":[{"count":0,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=\/wp\/v2\/posts\/2245057\/revisions"}],"wp:attachment":[{"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2245057"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2245057"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/analyse.optim.biz\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2245057"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}