Amathuluzi angu-1103Ikhasi 36 kwangu-46
![]()
WorkForgeI-ejenti ye-AII-WorkForge iwuhlaka lokwakha izindawo zokuqeqesha ama-agent oluqaliswe ngokubambisana yi-Tencent Hunyuan ne-Fudan University. I-WorkForge iqala ngohlaka lokugeleza komsebenzi lochwepheshe, ithole ngokuzenzakalelayo izinsiza zomhlaba wangempela ukuze yakhe indawo yokusebenza, bese ikhiqiza imisebenzi esekelwe emaqinisweni asendaweni yokusebenza...00![]()
MOSS-ImageUkuklama kwe-UI/UXI-MOSS-Image iyimodeli yokukhiqiza izithombe ngombhalo, ekhishwe ngomthombo ovulekile ngokubambisana yi-Mosi Intelligence ne-OpenMOSS. Inomgogodla we-DiT onamapharamitha angu-5B, ihlanganiswe ne-encoder yombhalo ye-Qwen3.5-9B efriziwe kanye ne-FLUX.2 VAE, kanti izisindo zayo zisebenzisa i-Apache-2.0.00![]()
MemT 1.0Amamodeli amakhulu e-AII-MemT 1.0 iyimodeli yokuqala eyisisekelo embonini eyakhelwe inkumbulo, eyethulwe yi-Memory Tensor futhi esivulwe ngokuphelele. I-MemT 1.0 inezinguqulo ezimbili: i-Flash (amapharamitha angu-288B/asebenzayo angu-17B) ne-Nano (amapharamitha angu-36B/asebenzayo angu-3B).00
BaoCutI-ejenti ye-AII-BaoCut iyithuluzi le-AI Video Agent elivulekile lomthombo elakhiwe ngu-Baoyu, elisekela ukuguqula inkulumo ibe umbhalo, ukuhumusha, ukwenza izwi, ukuhlela amavidiyo enkulumo, nokwengeza imiphumela yokunyakaza. Ekulungiseni ngakunye kushintshwa kuphela umusho noma isithombe esithile, ngaphandle kokuphinda ku-renderwe yonke ividiyo. Umuntu ne-Agent ba...00![]()
AgenshUkuhlela nge-AII-Agensh iwuhlaka lokusebenzisana kwama-agent amaningi olwakhiwe yi-Microsoft Research. Wonke ama-agent alandela umjikelezo ofanayo wezinyathelo ezinhlanu wokusebenzisana, ezihlela ngokwazo ukwaba umsebenzi zisebenzisa ikhosombe lekhodi elabiwe, imiyalezo esheshayo kanye nebhodi elihlelekile.00![]()
Claude DuetI-ejenti ye-AII-Claude Duet iyi-ejenti ye-AI yomuntu siqu eyethulwe yi-Anthropic, eyakhiwe ngemodeli ye-Opus 5.5. I-Claude Due ingaxhuma ngokuqhubekayo kuma-imeyili, ekhalendeni, emabhange nasezinhlelweni zamabhizinisi, ithumele isifinyezo somsebenzi njalo ekuseni, futhi ilungiselele izinto zokucubungula kusenesikhathi...00![]()
Fotor AgentI-ejenti ye-AII-Fotor Agent iyi-Agent ka-Fotor yokudala amavidiyo e-AI, esebenzisa indlela yobuchwepheshe ye-Video as Code. Abasebenzisi badinga kuphela ukuchaza izidingo zabo ngolimi lwemvelo, balayishe izithombe, amavidiyo, amadokhumenti noma izixhumanisi, bese i-AI ihlela ngokuzenzakalelayo okuqukethwe...00![]()
Eleven RealtimeI-ejenti ye-AII-Eleven Realtime iwuchungechunge lwamamodeli e-ejenti yezwi ahlanganisiwe olwenziwe yi-ElevenLabs, oluhlanganisa izinguqulo ezimbili: i-v1 Mini ne-v1 Max. I-Eleven Realtime ihlanganisa ukuqaphela inkulumo, ukucabanga kwemodeli yolimi, ukubiza amathuluzi kanye nok...00![]()
DeepInferAmamodeli amakhulu e-AII-DeepInfer iwuhlaka lokucabanga olusezingeni lokukhiqiza olukhishwe yi-DeepSeek njengomthombo ovulekile, olususelwa ku-vLLM, olusekela i-V4.1 Flash, futhi olusebenza ku-NVIDIA SM100 naku-Huawei Ascend 950. I-DeepInfer isebenzisa ukuthunyelwa okuhlukanisa i-PD, isebenzisa inqolobane ye-3FS...00![]()
Xing4.0-29B-IntentInkundla yokuthuthukisaI-Xing4.0-29B-Intent imodeli ekhethekile yokuqonda izinhloso ekhishwe njenge-open source yi-Xingchen Laboratory, eqeqeshwe kabusha kusetshenziswa i-Xing4.0-29B-A4B. I-Xing4.0-29B-Intent isebenzisa isakhiwo sesigaba＋amapharamitha, iguqula ulimi lwemvelo lomsebenzisi lube ngesikhathi esisodwa ukuhlukaniswa kwenhloso...00![]()
CleanYourTVUkuhlela nge-AII-CleanYourTV iyithuluzi lomthombo ovulekile lokuthuthukisa amathelevishini. Ixhuma ku-Android TV nge-ADB isebenzisa i-Wi-Fi, bese isebenzisa ama-AI Agent afana no-Claude Code ukuze ibone futhi ikhubaze ngokuzenzakalelayo izikhangiso, ukuqoqwa kwedatha kanye nezinhlelo ezifakwe kusengaphambili ezingadingekile, iphinde ifake...00![]()
Ming-3.0Ukwenziwa kwenkulumo nge-AII-Ming-3.0 iwuchungechunge lwamamodeli ezwi olukhishwe njenge-open source yi-Ant Group's Bailin, oluhlanganisa imisebenzi emithathu: ukuhlanganisa izwi, ukuqonda nomsindo nengxoxo, kanye nokubona izwi. Lethula amamodeli amahlanu: uchungechunge lwe-Ming-3.0-tts lusekela “Imodi Yomqondisi” yokulawula ngokuningiliziwe imizwa, ijubane...00![]()
FLUX 3 MaxUkuklama kwe-UI/UXI-FLUX 3 Max iyimodeli yevidiyo eqeqeshwe ngemuva eyethulwe ngokubambisana yi-fal Research ne-Black Forest Labs, esekela umbhalo ukuya kuvidiyo, isithombe ukuya kuvidiyo, kanye nokuhlanganiswa kozimele ababalulekile. Ingakhiqiza amavidiyo anomsindo wokuqala anobude bemizuzwana engu-5–20, nge-720p/1080p...00![]()
MiniMax M3.1 FlashI-ejenti ye-AII-MiniMax M3.1 Flash iyimodeli entsha yesizukulwane se-Flash ekhishwe yi-MiniMax, futhi iyimodeli engaziwa ye-'Space Bunny', ephatha ama-Token angaphezu kuka-200T. I-MiniMax M3.1 Flash isezingeni eliphezulu emisebenzini ye-Agentic, ekubhaleni izinhlelo, ku-Computer...00![]()
懂了鸭Inkundla yokufunda nge-AI懂了鸭 umphakathi wokufunda ojabulisayo osebenzisa i-AI, owethulwe yi-Tencent futhi ovulekele abasebenzisi abazali nezingane kanye nothisha abasebenza emgqeni wokuqala.懂了鸭 isebenzisa i-Qimiao Card njengendlela eyinhloko, iguqule okuqukethwe kolwazi okufana nezincwadi zezithombe, opopayi, izifundo eziguqukayo nemidlalo elula kube amafomu asebenzisanayo adonswayo, alingiswayo futhi asebenzisekayo...00![]()
EdenLiveUmphakathi we-AI yomthombo ovulekileI-EdenLive iyirenderer yokukhiqiza ngesikhathi sangempela eyethulwe yi-VAST, futhi iwumphumela wesigaba sephrojekthi yayo yemodeli yomhlaba i-Project Eden. I-EdenLive isebenzisa izigcawu ezilula ze-Proxy noma amavidiyo ayisibonelo njengamasignali okulawula, ihlanganise nohlaka lokuqala oluqondiwe kanye...00![]()
ScholarMuseInkundla yokuthuthukisaI-ScholarMuse iyinkundla yamathuluzi e-MCP eyenzelwe abacwaningi, exhuma imithombo yedatha yocwaningo engaphezu kuka-50 ku-Claude, ChatGPT, Codex, Cursor nabanye abasizi be-AI asebevele besetshenziswa.00![]()
CiteRealUkubhala nge-AII-CiteReal iyithuluzi le-AI lokuqinisekisa izingcaphuno nokulandela imithombo yezincwadi, elisiza ababhali bahlole ukuthi izinkomba zabo ziyiqiniso yini futhi ziyatholakala. I-CiteReal inikeza imisebenzi emibili: ukuqinisekisa izingcaphuno nokuthola imithombo. Eyokuqala iqhathanisa isihloko, ababhali, unyaka, i-DOI...00![]()
VoidInkundla yokuthuthukisaIVoid iyithuluzi elivulekile lokuthuthukisa nokuphakela ama-full-stack elakhiwe yiqembu lika尤雨溪, elihlanganisa izingxenye ezintathu: i-SDK, i-CLI kanye nenkundla yokuphakela amaqembu. IVoid ihlanganisa ukubhala ikhodi yohlelo lwe-Vite, ukuxhuma isizindalwazi nokulufaka ku-inthanethi kube yinqubo eyodwa: ukuthuthukiswa kwasendaweni kulungiselela ngokuzenzakalelayo izinsiza zendawo...00![]()
SynvasUkuhlela nge-AII-Synvas iyithuluzi lokudala amavidiyo e-AI elisekelwe kukhanvasi, elakhiwe ngokubambisana ngu-Ding Zhiyuan (Julian), usomabhizinisi osungula amabhizinisi njalo, kanye nomklami u-WZ. I-Synvas isebenzisa ubuchwepheshe bezinjini zemidlalo (WebGPU + WebCodecs), ibiza ngokuqondile i-GPU yasendaweni,...00
Hy Memory 2.0Inkundla yokuthuthukisaI-Hy Memory 2.0 iwuhlaka lwememori lwama-agent olukhishwe yi-Tencent Hunyuan njengomthombo ovulekile, olusebenzisa ukwakheka kwezingqimba ezintathu: imodeli + uhlaka + uhlelo lokusebenza: imodeli i-Hy-Memory-30A3B eqeqeshelwe inkumbulo, i-SDK esekela izinkumbulo zengxoxo/umsebenzi/ikhodi, kanye...00![]()
BlenderWorldUkuhlolwa kwamamodeli e-AII-BlenderWorld iyibhentshimakhi yokuhlola ama-Agent okudala i-3D eyethulwe ngokubambisana yiQwen team, iZhejiang University, iChinese University of Hong Kong, iFudan University kanye neNational University of Singapore. Iqukethe imisebenzi yangempela engu-223, futhi ihlola ukuthi i-AI ingakwazi yini, isebenzisa izithombe, amavidiyo noma imininingwane yokuklama, ku...00
DeepSeek-V4.1-Flash-1017Ukuhlela nge-AII-DeepSeek-V4.1-Flash-1017 inguqulo ethuthukisiwe yemodeli elula ekhishwe yi-DeepSeek; iwebhu, i-APP ne-API sekushintshelwe kuyo kanyekanye. Inguqulo entsha iqeqeshwe ngokufunda okuqinisiwe okusekelwe ekuzihloleni, ithuthukisa kakhulu amakhono okuthuthukisa imidlalo, ukumodela kwe-3D nokunye...00
MiDashengLM-SpatialAmamodeli amakhulu e-AII-MiDashengLM-Spatial iyimodeli enkulu yokuqonda umsindo wendawo evulelwe umphakathi yithimba lakwa-Xiaomi Dasheng, futhi iyimodeli yokuqala embonini ehlanganisa ukuqonda umsindo okuvamile nokuqonda indawo ngaphakathi kwesakhiwo esisodwa, kusukela ekuqaleni kuze kube sekugcineni.00
