{"id":666835,"date":"2025-12-17T18:16:04","date_gmt":"2025-12-17T18:16:04","guid":{"rendered":"https:\/\/microscopemedia.com\/?p=666835"},"modified":"2025-12-17T18:16:04","modified_gmt":"2025-12-17T18:16:04","slug":"google-lanseaza-modelul-rapid-si-accesibil-gemini-3-flash-concurand-direct-cu-openai","status":"publish","type":"post","link":"https:\/\/microscopemedia.com\/?p=666835","title":{"rendered":"Google lanseaz\u0103 modelul rapid \u0219i accesibil Gemini 3 Flash, concur\u00e2nd direct cu OpenAI"},"content":{"rendered":"
<\/div>\n

Noul model vine la doar o lun\u0103 dup\u0103 Gemini 3 \u0219i la \u0219ase luni de la Gemini 2.5 Flash, oferind performan\u021be semnificativ \u00eembun\u0103t\u0103\u021bite \u0219i viz\u00e2nd s\u0103 concureze cu modelele de frontier\u0103, inclusiv GPT-5.2 de la OpenAI, scrie Tech Crunch<\/a>.<\/p>\n

Pe benchmark-ul Humanity\u2019s Last Exam, proiectat pentru a testa expertiza pe diverse domenii, Gemini 3 Flash a ob\u021binut 33,7%, dep\u0103\u0219ind Gemini 2.5 Flash care a \u00eenregistrat doar 11%, \u0219i apropiindu-se de GPT-5.2 (34,5%) \u0219i Gemini 3 Pro (37,5%). \u00cen testul multimodal \u0219i de ra\u021bionament MMMU-Pro, modelul a \u00eenregistrat un scor impresionant de 81,2%, dep\u0103\u0219ind toate modelele concurente.<\/p>\n

Func\u021bionalit\u0103\u021bi \u0219i disponibilitate pentru consumatori<\/h2>\n

Modelul Gemini 3 Flash<\/strong> devine implicit \u00een aplica\u021bia Gemini la nivel global, \u00eenlocuind versiunea 2.5 Flash. Utilizatorii pot alege \u00een continuare modelul Pro pentru \u00eentreb\u0103ri mai complexe de matematic\u0103 \u0219i programare. Noua versiune se remarc\u0103 prin capacitatea de a interpreta con\u021binut multimodal: utilizatorii pot \u00eenc\u0103rca videoclipuri scurte, desene sau fi\u0219iere audio pentru analize detaliate, sugestii sau generarea de teste.<\/p>\n

Modelul ofer\u0103 r\u0103spunsuri mai vizuale, incluz\u00e2nd imagini \u0219i tabele, \u0219i poate fi folosit pentru prototiparea de aplica\u021bii direct \u00een Gemini. Google sus\u021bine c\u0103 Gemini 3 Flash este mai rapid \u0219i mai eficient dec\u00e2t predecesorul s\u0103u, folosind cu 30% mai pu\u021bine token-uri \u00een sarcinile de g\u00e2ndire complex\u0103 \u0219i oferind o alternativ\u0103 mai economic\u0103 pentru sarcinile repetitive.<\/p>\n

Disponibilitate pentru companii \u0219i dezvoltatori<\/h2>\n

Companii precum JetBrains, Figma, Cursor, Harvey \u0219i Latitude folosesc deja Gemini 3 Flash prin Vertex AI \u0219i Gemini Enterprise. Pentru dezvoltatori, modelul este disponibil \u00een preview prin API \u0219i instrumentul de programare Antigravity. Pre\u021bul este de 0,50 dolari per 1 milion token-uri de intrare \u0219i 3,00 dolari per 1 milion token-uri de ie\u0219ire, oferind un raport cost-eficien\u021b\u0103 avantajos comparativ cu Gemini 2.5 Flash.<\/p>\n

Google raporteaz\u0103 c\u0103, de la lansarea Gemini 3, se proceseaz\u0103 peste 1 trilion de token-uri zilnic pe API, \u00eentr-o competi\u021bie str\u00e2ns\u0103 cu OpenAI, care a lansat recent GPT-5.2 \u0219i un nou model de generare a imaginilor.<\/p>\n

Tulsee Doshi, director senior Gemini Models, a declarat c\u0103 lans\u0103rile recente \u201cprovoc\u0103 industria s\u0103 inoveze \u0219i s\u0103 creeze modele din ce \u00een ce mai performante\u201d, \u00eencuraj\u00e2nd adoptarea unor noi benchmark-uri pentru evaluarea AI.<\/p>\n","protected":false},"excerpt":{"rendered":"

Noul model vine la doar o lun\u0103 dup\u0103 Gemini 3 \u0219i la \u0219ase luni de la Gemini 2.5 Flash, oferind performan\u021be semnificativ \u00eembun\u0103t\u0103\u021bite \u0219i viz\u00e2nd s\u0103 concureze cu modelele de … Read More<\/a><\/p>\n","protected":false},"author":1,"featured_media":666836,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"Default","format":"standard","meta":[],"categories":[1],"tags":[],"_links":{"self":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/posts\/666835"}],"collection":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=666835"}],"version-history":[{"count":0,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/posts\/666835\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=\/wp\/v2\/media\/666836"}],"wp:attachment":[{"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=666835"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=666835"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/microscopemedia.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=666835"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}