AMABWIRIZA
Tencent Hunyuan
Hunyuan, ubu akenshi yitwa Hy, ni umuryango wa Tencent w'ibyitegererezo fatizo, urimo ibyitegererezo binini by'indimi, ugukora amashusho ndetse no gukora ibintu bya 3D.
Umva iki gitaboAI-byavuzwe · ~8 min
Kuriyi page9 min soma
Incamake
Yafunguwe bwa mbere mu bucuruzi muri Nzeri 2023, iha imbaraga ibicuruzwa bya Tencent nkumufasha wa Yuanbao, hamwe n’ibisohoka byashyizwe ahagaragara, harimo miliyari 770-ibipimo bya Hy4 byerekanwe muri Kanama 2026, bituma iba imwe mu miryango y’icyitegererezo y’Ubushinwa.
Ibyingenzi byingenzi
- Hunyuan, yiswe Hy, ni umuryango wa Tencent wururimi, amashusho na 3D fondasiyo ya 3D, yafunguwe bwa mbere mubucuruzi ku ya 7 Nzeri 2023.
- Hunyuan-Kinini, yasohotse mu Gushyingo 2024, ni imvange-y’impuguke ifite miliyari 389 zose hamwe na miliyari 52 zikora.
- Hy3 ibanziriza, ifunguye-isoko ku ya 23 Mata 2026, ifite miliyari 295 zose hamwe, miliyari 21 zikora hamwe nidirishya rya 256.000.
- Hy4 ibanziriza, yasohotse ku ya 28 Kanama 2026 munsi ya Apache 2.0, ifite miliyari 770 zose hamwe, miliyari 49 zikora hamwe na miliyoni zirenga zerekana imiterere.
- Imibare myinshi yatangajwe Hunyuan yonyine yakozwe na Tencent kandi ntabwo yagenzuwe yigenga.
Kwibira cyane
Hunyuan is the brand Tencent, the Chinese technology company that owns WeChat, uses for its in-house foundation models. A foundation model is a large neural network trained on broad data that can then be adapted to many tasks. The Hunyuan family is not one model but several lines: general large language models for text and code, image generation models, the HunyuanVideo line for video and the Hunyuan3D line for 3D assets. Recent model cards and announcements use the shorter name Hy, as in Hy3 and Hy4, and describe the developer as the Tencent Hy team. The models serve two audiences: Tencent's own products, such as the Yuanbao chatbot and the CodeBuddy and WorkBuddy assistants, and outside developers through Tencent Cloud and open-weight downloads.
The first public milestone came on 7 September 2023, when Tencent demonstrated Hunyuan in Shenzhen and opened it to enterprise customers. Reuters reported that Tencent described the model as having more than 100 billion parameters, the adjustable values a network learns in training, and as trained on more than 2 trillion tokens, the word fragments a language model reads. Tencent claimed Hunyuan hallucinated, meaning produced confident false statements, 30 percent less than Meta's Llama 2, and said it already underpinned more than 50 internal products; Reuters noted these claims were not independently verified. The launch came in a crowded field: Tencent vice president Jiang Jie said China already had more than 130 large language models by July 2023.
In November 2024 Tencent made its most significant early open release, Hunyuan-Large. Its arXiv paper describes a mixture-of-experts (MoE) model with 389 billion total parameters but only 52 billion activated for each token. In an MoE model, each layer contains many smaller expert sub-networks and a router picks a few of them for each token, so the model can store a lot of knowledge while spending compute like a much smaller model. The paper reports a context window of up to 256,000 tokens, heavy use of synthetic training data and compression of the key-value cache, the memory a transformer keeps about earlier tokens, to make long inputs cheaper. Code and checkpoints were released publicly.
Tencent also built a strong presence in open visual generation. HunyuanVideo, posted in December 2024, is a video generation model with over 13 billion parameters, which its authors described as the largest open-source model of its kind at the time; in professional human evaluations they report it outperformed Runway Gen-3, Luma 1.6 and three leading Chinese video models. Hunyuan3D 2.0, posted in January 2025, splits 3D generation into two models: Hunyuan3D-DiT, a flow-based diffusion transformer that generates a bare shape matching a condition image, and Hunyuan3D-Paint, which generates high-resolution texture maps for that mesh or for one made by hand. A diffusion transformer is an image or shape generator that starts from noise and removes it step by step under the control of a transformer network. The authors report it beats earlier open and closed systems on geometry detail and texture quality, though these are self-reported results.
The language-model line was then rebuilt. Hy3 preview, open-sourced on 23 April 2026, is described on its model card as the first model trained on the team's rebuilt infrastructure. Its model card lists 295 billion total parameters with 21 billion active, 192 experts of which 8 are active per token, a 256,000-token context window and a choice of reasoning effort from none to high. It reports 87.2 on the GPQA Diamond science benchmark and 74.4 percent of tasks resolved on SWE-bench Verified, a test of fixing real software issues. The final Hy3 launched on 6 July 2026, with more post-training compute and better data according to Caixin, which also reported Tencent's claim of a 90 percent completion rate on agent tasks in internal applications.
Hy4 preview followed on 28 August 2026. Tencent's announcement lists 770 billion total parameters, 49 billion active and a context window above one million tokens, with API prices of about 0.83 US dollars per million input tokens and 2.50 dollars per million output tokens. The Hugging Face model card describes 78 layers, 256 routed experts plus one shared expert with the top 8 routed experts used per token, and an Apache 2.0 license, a permissive open-source license that allows commercial use. It reports scores such as 92.3 on GPQA Diamond and 65.7 on SWE-bench Pro. In an internal blind test of 203 engineering tasks judged by 163 experts, Tencent reported an average rating of 2.99 out of 4, against 2.92 for GLM-5.3 and 2.94 for Kimi K3.
Compared with other Chinese model families, including GLM and Kimi, which Tencent itself uses as reference points, Hunyuan stands out less for a single architectural idea than for breadth: Tencent maintains competitive language, video and 3D lines at the same time and ties them to products with very large user bases. Its recent language models follow the same broad recipe as many peers, large sparse MoE transformers with long context and adjustable reasoning, and the Hy4 card names DeepSeek Sparse Attention as part of its design. Licensing has varied by release, which matters for builders: Hy3 preview uses the Tencent Hy Community License, while Hy4 preview uses Apache 2.0.
The main limitations are about verification and access. Most headline numbers, from the 2023 hallucination claim to the Hy4 blind evaluation, come from Tencent itself, and the comparison models are chosen by Tencent. Benchmarks such as GPQA and SWE-bench measure narrow skills and can be affected by training data overlap. Running the largest open models is expensive: Tencent's own deployment examples for Hy3 and Hy4 use eight large-memory GPUs, and 295 billion parameters stored in 16-bit precision alone take roughly 590 GB. For most users, practical access is through Tencent's apps or paid APIs rather than self-hosting.
04Urugero rwakazi
Soma ikarita yicyitegererezo ya Hunyuan nka injeniyeri
Fungura ikarita yicyitegererezo ya Hy3 kuri Hugging Face hanyuma wandike ibipimo byose, ibipimo bikora nuburebure bwimiterere.
Gabanya ibikorwa ukurikije ibipimo byose kugirango ugereranye igice cyurusobe rukora kuri buri kimenyetso.
Kugwiza ibipimo byose kuri 2 bytes kugirango ugereranye kwibuka kuburemere bwa 16-bit, mbere yingenzi-agaciro cache.
Reba ahabigenewe uruhushya hanyuma usome niba byemerera gukoresha ubucuruzi bwawe.
Subiramo ikarita ya Hy4-ibanziriza hanyuma ugereranye kubara impuguke, idirishya ryuruhushya.
Andika amanota y'ibipimo biva mu isuzuma rya Tencent wenyine kandi niba bihari, biva mubuyobozi bwigenga.
Ibyo byerekana
Abiga barangiza bagereranya urupapuro rumwe rwerekana ko Hy3 ibanziriza ikora hafi 7 ku ijana yibipimo bya miliyari 295 kuri buri kimenyetso kandi ikenera hafi 590 GB kuburemere gusa, hamwe nimpushya zerekana ibipimo bya buri gusohora. Imibare irashobora kugenzurwa neza namakarita yicyitegererezo.
Gushyira mu bikorwa Isi
Tencent yavuze muri Nzeri 2023 ko Hunyuan yamaze kuba umusingi w’ibicuruzwa na serivisi birenga 50 igihe yafunguriraga icyitegererezo abakiriya bo mu bigo byo hanze.
HunyuanVideo, yasohowe hamwe na kode ifunguye mu Kuboza 2024, ni icyitegererezo cyanditse kuri videwo gifite ibipimo bisaga miliyari 13 abanditsi bacyo bavuze hejuru ya Runway Gen-3 na Luma 1.6 mu isuzuma ry’umwuga.
Hunyuan3D 2.0 ihindura ishusho imwe mesh ya 3D mesh mubyiciro bibiri, imiterere hanyuma imiterere, hamwe na code hamwe nuburemere byatangajwe kuri GitHub kumikino no gushushanya.
Hy3, icyitegererezo cy’ururimi rwo muri Nyakanga 2026, yinjijwe mu mufasha w’abaguzi ba Tencent Yuanbao ku munsi wo gutangiza, Caixin avuga ko Tencent yagabanije amafaranga yo gukoresha abaterankunga icyarimwe.
Abashoramari barashobora gukoresha Hy4 ibanziriza ibipimo bifunguye hamwe na vLLM cyangwa SGLang bakoresheje inzira-8 ya tensor parallelism, cyangwa bakayita binyuze muri Tencent Cloud TokenHub na OpenRouter.
Kazoza ka Tencent Hunyuan
Tencent's own statements point to more frequent releases with a preview-first pattern: its August 2026 Hy4 announcement says the team publishes previews, gathers real-world feedback and then ships official releases, and that more Hy4-series models were expected soon. The same announcement emphasizes agent work, such as coding, office documents and game prototyping, and claims Hy4 preview took part in automated optimization of its own training methods and inference system, a claim that has not been independently assessed. Pricing is also moving: Caixin reported that Tencent cut developer fees when the final Hy3 launched in July 2026, and the Hy4 preview announcement offered two weeks of free use in CodeBuddy and WorkBuddy at launch. Open questions include how Hunyuan models perform on independent evaluations rather than Tencent-selected comparisons, whether the licensing shift to Apache 2.0 with Hy4 preview continues for future releases, and how the visual lines, HunyuanVideo and Hunyuan3D, are integrated with the language models as Tencent moves toward multimodal agents.
Kubishyira mubikorwa
Impamvu ari ngombwa
- Ingamba z'abacuruzi. Ibishushanyo mbonera byabacuruzi bigira ingaruka kubiranga ikipe yawe ishobora kubaka ubutaha.
- Igiciro na bije. Amagambo yubucuruzi nuburyo bwo kohereza bigira ingaruka kubiciro byigihe kirekire ningaruka.
- Ibyago n'umutekano. Isosiyete ishimangira gushiraho ibicuruzwa bitemewe, igihagararo cyumutekano, no gufungura.
Witondere
- Gutangiza amatangazo arashobora gusumbya ituze mubikorwa nyabyo byakazi.
- Ibiciro bya API cyangwa guhindura politiki birashobora guhagarika ibitekerezo ijoro ryose.
- Abashoramari bonyine bishingira byongera gufunga no kwimuka.
Intambwe ikurikira
- Suzuma abatanga serivisi ukoresheje imirimo yawe bwite na datasets.
- Ongera usuzume ubuzima bwite, umutekano, namategeko mbere yo kwishyira hamwe.
- Komeza gahunda yo gusubira inyuma kurugero cyangwa abacuruzi.
- Kurikirana inyandiko zisohora kugirango impinduka zumuhanda ntizitangaje amakipe.
Inkomoko no gusoma
- Reuters via MalaysiaNowUbushinwa Tencent bwatangije imvugo nini ya AI, ivuga ko ifunguye gukoresha imishinga
- Hugging Face (Tencent Hy team)tencent/Hy4-preview
- arXiv (Tencent)Hunyuan-Kinini: Gufungura-isoko ya MoE Model hamwe na miliyari 52 zikoreshwa na Tencent
- arXiv (Tencent)HunyuanVideo: Urwego rutunganijwe kuri Video nini yerekana amashusho
- arXiv (Tencent)Hunyuan3D 2.0
- Hugging Face (Tencent Hy team)tencent/Hy3-preview
- Caixin GlobalTencent yatangije Hunyuan 3 AI yerekana imiterere yubuntu
- TencentTencent Isohora hamwe na Gufungura-Inkomoko Tencent Hy4 ibanziriza
Komeza Ubushakashatsi
Ibaruwa y'amakuru y'ubuntu
Ibuka amakuru ya buri munsi ya AI
Inkuru eshatu zizewe za AI buri gitondo cy'iminsi y'akazi, zanditse mu rurimi rworoshye. Ubuntu iteka ryose, nta kwamamaza.
Imeri imwe buri munsi y'akazi. Kwiyimura mu buryo bumwe. Ntitwigera tugurisha cyangwa ngo dusangire aderesi yawe.
Isuzume
Fata ikibazo cya Tencent Hunyuan
Ibisubizo ako kanya kuri buri kibazo, n'icyemezo gishobora gusangirwa gifite ID yagenzurwa umaze gutsinda isomo.
Fasha uburezi bwa AI bw'ubuntu. AI Understanding ni umuryango utegamiye kuri leta wa 501(c)(3) — nta kwamamaza, nta kwishyura, iteka ryose. Tanga impano
Ibibazo bikunze kubazwa
Tencent Hunyuan ni iki?
Hunyuan, ubu ikunze kugabanywa na Hy, ni umuryango wa Tencent wintangarugero zifatizo, zikubiyemo imiterere nini yindimi, kubyara amashusho no kubyara umutungo wa 3D. Yafunguwe bwa mbere mu bucuruzi muri Nzeri 2023, iha imbaraga ibicuruzwa bya Tencent nkumufasha wa Yuanbao, hamwe n’ibisohoka byashyizwe ahagaragara, harimo miliyari 770-ibipimo bya Hy4 byerekanwe muri Kanama 2026, bituma iba imwe mu miryango y’icyitegererezo y’Ubushinwa.
Niki Hunyuan3D yagenewe kubyara?
Hunyuan3D ibyara umutungo wa 3D uhereye kumyandiko cyangwa amashusho yinjiza, afite akamaro mumikino no gushushanya.
Nibihe bicuruzwa bya Tencent umufasha wa AI ukoreshwa na Hunyuan?
Tencent Yuanbao numufasha wumuguzi AI wubatswe kuri moderi ya Hunyuan.
Ni ubuhe bwoko bw'icyitegererezo Hunyuan-DiT?
Urupapuro rwa Hunyuan-DiT rwa Tencent rusobanura impinduka nyinshi zo gukwirakwiza impinduka zo guhindura inyandiko-ku-shusho.
Niki Hunyuan3D 2.0 itanga mbere yicyiciro cyayo-synthesis?
Urupapuro rwa Hunyuan3D 2.0 rusobanura ibisekuru bya geometrie bikurikirwa na synthesis.
Ni ubuhe buryo Tencent yakoresheje kugira ngo Hunyuan yemerwe?
Tencent yafunguye-amasoko menshi ya Hunyuan, harimo 3D nindimi ntoya, kugirango ikurura abitezimbere.
Komeza wige
Kuyobora
Abandi bayobozi batoranijwe kuriyi ngingo