Tencent HY

653 posts

Tencent HY banner
Tencent HY

Tencent HY

@TencentHunyuan

Tencent's foundation model for text, image, video, and 3D generation.

Latent space Katılım Temmuz 2024
11 Takip Edilen34.7K Takipçiler
Sabitlenmiş Tweet
Tencent HY
Tencent HY@TencentHunyuan·
🌍Tencent HY 3D Studio is now live globally. An AI-powered workspace for professional 3D production pipelines. 🧱Image → 3D → topology → UVs → texturing → rigging → animation All in one workflow. Built for creators, game developers, and 3D artists. 🕹️ Try it: 3d.hunyuanglobal.com/studio New users get 20 free generations daily (limited-time). ⚙️ API access: tencentcloud.com/products/ai3d New developers receive 200 free credits to get started.
Tencent HY tweet mediaTencent HY tweet media
Tencent HY@TencentHunyuan

We are officially taking Hunyuan 3D Engine global! 🌎 Responding to community demand and building on open-source acclaim, Hunyuan 3D capabilities are now launching globally to empower all creators and enterprises! This next-gen AI platform cuts commercial-grade 3D asset production from days or weeks down to minutes. 🎨Multimodal Input: Instantly generates from Text-to-3D, Image-to-3D (multi-view support), or Sketch-to-3D. 💡Pro-Ready Quality: Achieves industry SOTA with our new 3D-DiT hierarchical carving model, delivering a 3x boost in modeling precision (up to 15363 ultra-HD resolution). Seamless OBJ/GLB integration with Unreal, Unity, and Blender is standard. ☁️Commercial Access: The Hunyuan 3D model API is now available on Tencent Cloud International, enabling global enterprises to integrate advanced 3D generation into workflows for game development, e-commerce, advertising, 3D printing, and more. New creators get 20 free generations daily on the platform. Enterprise API users receive 200 free credits upon registration. Start building now! 👉 Try the creation engine: 3d.hunyuanglobal.com 🔗 Access the API: tencentcloud.com/products/ai3d

English
26
94
943
157.1K
Tencent HY
Tencent HY@TencentHunyuan·
We're releasing HY-Embodied-0.5, a family of foundation models for real-world embodied agents. The 2B model is now open source. It strengthens spatial-temporal perception and embodied reasoning for prediction, interaction, and planning. 🤖 The suite includes: 🔹 2B for edge deployment 🔹 32B for complex reasoning Key innovations: 🔹 Mixture-of-Transformers (MoT) architecture for modality-specific computation 🔹 Latent tokens for improved perceptual representation 🔹 Self-evolving post-training 🔹 On-policy distillation from large to small models Across 22 benchmarks, the 2B model outperforms similarly sized SOTA systems on 16 tasks. The 32B model approaches frontier-level performance. 🔗 GitHub: github.com/Tencent-Hunyua… 🤗 Hugging Face: huggingface.co/tencent/HY-Emb…
English
10
52
314
31.1K
Tencent HY retweetledi
GMI Cloud
GMI Cloud@gmi_cloud·
Hunyuan Image-to-Image is now supported on GMI and GMI Studio. Our first Hunyuan series model is officially here. @TencentHunyuan We’re building out the collection, so comment below with which other Hunyuan models you want to see on GMI next. (FYI: 3D is coming.)
English
2
5
31
3.9K
Tencent HY
Tencent HY@TencentHunyuan·
New research with @Tsinghua_Uni: Spatial-TTT. A framework for streaming visual-based spatial intelligence with test-time training (TTT). Spatial-TTT adapts fast weights to capture and organize spatial evidence from long video streams, enabling models to build structured 3D spatial memory over time. Highlights: 🔹Efficient streaming memory. Fast weights act as compact spatial memory with sublinear memory growth over 7000+ frames and more than 40% lower compute. 🔹Spatial-predictive mechanism. TTT layers with 3D spatiotemporal convolution capture geometric correspondence and temporal continuity. 🔹SOTA results on long-horizon video spatial understanding (VSI-Bench). The paper ranked #1 on @huggingface Daily Papers on March 13. Project page: liuff19.github.io/Spatial-TTT/ GitHub: github.com/THU-SI/Spatial… Paper: huggingface.co/papers/2603.12… Model & Data: huggingface.co/THU-SI
English
10
36
214
19.8K
Tencent HY
Tencent HY@TencentHunyuan·
We are open-sourcing WorldCompass, an RL post-training framework specifically designed for Interactive World Models. 🛠️ Open Training Code: Fully customizable for post-training with your own data, rewards, or base models. ⚡ Open-source Checkpoint: More precise instruction-following for complex, compositional action inputs. WorldCompass significantly improves compositional action following and long-horizon interaction in HY-World 1.5. 🕹️ Play now: 3d.hunyuan.tencent.com/sceneTo3D?tab=… ⭐ GitHub: github.com/Tencent-Hunyua…
Tencent HY tweet mediaTencent HY tweet media
English
2
22
187
15.3K
Tencent HY
Tencent HY@TencentHunyuan·
🏆HY-WU surpasses or rivals leading open- and closed-source models in human evaluation. (🧵5/6)
Tencent HY tweet media
English
1
1
15
3.9K
Tencent HY
Tencent HY@TencentHunyuan·
One static model does not fit all😭 We just dropped our latest work: Functional Neural Memory. Instead of static models, we generate custom "parameters" for every single input. ✅Prompt your model anytime ✅Instant personalization ✅Better instruction following ✅Flexible & dynamic memory (w/o memory bank✌️) (🧵1/6)
English
11
138
331
69.7K
Tencent HY
Tencent HY@TencentHunyuan·
@OdinLovis This is awesome! Love seeing creators extend the ecosystem like this. The WebGL version looks great😃
English
0
0
2
113
Lovis Odin
Lovis Odin@OdinLovis·
@TencentHunyuan Thank you for sharing !! and this is the "final" version available in webgl x.com/i/status/20157…
Lovis Odin@OdinLovis

"Hand-drawn sketch → full 3D WebGL experience ✨ (Made with Claude Code and @fal ) Play here 👇 pharos.graphics 60% my craft: drawing, C4D, Three.js shaders, baking 20% Claude Code 20% fal.ai: Image model + Hunyuan 3D AI doesn't replace creativity, it amplifies it. #threejs #WebGL #creativecoding #fal #ClaudeCode

English
1
0
6
647
Tencent HY
Tencent HY@TencentHunyuan·
Excited to see our 3D model integrated into creative workflows!🥰
Lovis Odin@OdinLovis

Old Art to full 3D using Cinema 4D . Created a free and open-source plugin to bring Hunyuan 3D directly into Cinema 4D. github.com/lovisdotio/cin… This workflow combines multiple tools that are available today into a single, cohesive pipeline. I wanted to show that AI is never an end product by itself, it’s not the finality. Used properly, AI becomes an extremely powerful tool that lets you build exactly what you want. I hope you like it. The final step will be to bake all textures so I can share the project as a WebGL experience, a small interactive website to explore it directly in the browser. Made with heart, and with @fal ❤️

English
2
10
149
13.3K
Tencent HY retweetledi
Hirokazu Yokohara
Hirokazu Yokohara@Yokohara_h·
画像から3D生成 いつの間にかHunyuan3Dが3.1にアップデートされていたのでテストしてみた 以前よりメッシュのディティールが奇麗になってて素晴らしい
日本語
17
692
7.1K
1.1M
Tencent HY
Tencent HY@TencentHunyuan·
More than 25 papers from Tencent HY team have been accepted to @CVPR 2026. This year, our work spans: 🖼️ Image Editing: Chain-of-Thought (CoT) reasoning and tool-calling for fine-grained, controllable editing. 🎥 Video Generation: Post-training RL, unified audio-video generation, streaming video generation, and inference acceleration. 🎮 3D Vision: Controllable 3D generation, texture synthesis, 3D reinforcement learning, interactive world models, and unified multimodal 3D models. From reasoning-guided editing to scalable video and 3D-native foundations, we focus on building general generative models across modalities. We’ll be sharing technical deep dives during the conference. See you at #CVPR2026 in Denver this June!
Tencent HY tweet mediaTencent HY tweet mediaTencent HY tweet mediaTencent HY tweet media
English
4
24
233
15.7K