NVIDIA AI Developer's banner
NVIDIA AI Developer's profile picture

NVIDIA AI Developer

@NVIDIAAIDev • 111,125 subscribers

This account is inactive. Follow @NVIDIAAI for the latest updates.

Shorts

Today, we released Lyra 2.0, a framework for generating persistent, explorable 3D worlds at scale, from NVIDIA Research. Generating large-scale, complex environments is difficult for AI models. Current models often “forget” what spaces look like and lose track of movement over time, causing objects to shift, blur, or appear inconsistent. This prevents them from creating the reliable 3D environments required for downstream simulations. Lyra 2.0 solves these issues by: ✅ Maintaining per-frame 3D geometry to retrieve past frames and establish spatial correspondences ✅ Using self-augmented training to correct its own temporal drifting. Lyra 2.0 turns an image into a 3D world you can walk through, look back, and drop a robot into for real-time rendering, simulation, and immersive applications. ➡️ Learn more: 📄 Read the paper:

Today, we released Lyra 2.0, a framework for generating persistent, explorable 3D worlds at scale, from NVIDIA Research. Generating large-scale, complex environments is difficult for AI models. Current models often “forget” what spaces look like and lose track of movement over time, causing objects to shift, blur, or appear inconsistent. This prevents them from creating the reliable 3D environments required for downstream simulations. Lyra 2.0 solves these issues by: ✅ Maintaining per-frame 3D geometry to retrieve past frames and establish spatial correspondences ✅ Using self-augmented training to correct its own temporal drifting. Lyra 2.0 turns an image into a 3D world you can walk through, look back, and drop a robot into for real-time rendering, simulation, and immersive applications. ➡️ Learn more: 📄 Read the paper:

432,726 views

Most 3DGS segmentation tools either pre‑train per scene or lock errors into a feature field you can’t undo. ArtisanGS instead turns a few 2D masks into editable 3D object selections via Cutie tracking + black‑box splat aggregation, then lets you iteratively correct mistakes with consistent 2D/3D selection modes. 📗 #NVIDIAResearch paper:

Most 3DGS segmentation tools either pre‑train per scene or lock errors into a feature field you can’t undo. ArtisanGS instead turns a few 2D masks into editable 3D object selections via Cutie tracking + black‑box splat aggregation, then lets you iteratively correct mistakes with consistent 2D/3D selection modes. 📗 #NVIDIAResearch paper:

42,107 views

🤖 One predictive backbone, three distinct tasks, consistent gains: a strong signal that investing in reusable world models is the right abstraction for robot learning. Toyota Research Institute (TRI) just ran the same idea three different ways—and it worked each time. Using our NVIDIA Cosmos Predict 2–style world models, they hit SOTA with dynamic view synthesis, teleop data augmentation (even up against MimicGen as an upper bound), and navigation world models. 🙌 📚 dynamic view synthesis 📚 teleop data augmentation (also compare to MimicGen as an upper bound) 📚 navigation world models 📸 from

🤖 One predictive backbone, three distinct tasks, consistent gains: a strong signal that investing in reusable world models is the right abstraction for robot learning. Toyota Research Institute (TRI) just ran the same idea three different ways—and it worked each time. Using our NVIDIA Cosmos Predict 2–style world models, they hit SOTA with dynamic view synthesis, teleop data augmentation (even up against MimicGen as an upper bound), and navigation world models. 🙌 📚 dynamic view synthesis 📚 teleop data augmentation (also compare to MimicGen as an upper bound) 📚 navigation world models 📸 from

38,100 views

A preview Into the future of #quantum-GPU computing: At GTC Washington, D.C., we announced NVIDIA NVQLink, an open system architecture for tightly coupling the extreme performance of GPU computing with quantum processors to build accelerated quantum supercomputers. In the expo hall, accelerated quantum computing was a centerpiece of the NVIDIA booth. See our live blog:

A preview Into the future of #quantum-GPU computing: At GTC Washington, D.C., we announced NVIDIA NVQLink, an open system architecture for tightly coupling the extreme performance of GPU computing with quantum processors to build accelerated quantum supercomputers. In the expo hall, accelerated quantum computing was a centerpiece of the NVIDIA booth. See our live blog:

54,981 views

While at #OpenSourceAIWeek, Andrej Karpathy received one of our first NVIDIA DGX Sparks, personally delivered by Nader Khalil🍊 and @Baxate_carter. 🙌 Excellent work by Andrej on introducing Nanochat -- an #opensource project that makes it easy for developers to explore and train their own LLMs from scratch. We can’t wait to see what you’ll do with your Spark! #SparkSomethingBig ✨

While at #OpenSourceAIWeek, Andrej Karpathy received one of our first NVIDIA DGX Sparks, personally delivered by Nader Khalil🍊 and @Baxate_carter. 🙌 Excellent work by Andrej on introducing Nanochat -- an #opensource project that makes it easy for developers to explore and train their own LLMs from scratch. We can’t wait to see what you’ll do with your Spark! #SparkSomethingBig ✨

48,804 views

🎉 Announcing NVIDIA DGX Spark (f.k.a. Project DIGITS). Are you ready to #SparkSomethingBig? 💫 ➡️ #GTC25

🎉 Announcing NVIDIA DGX Spark (f.k.a. Project DIGITS). Are you ready to #SparkSomethingBig? 💫 ➡️ #GTC25

58,868 views

Brain-Computer Interfaces for speech just overcame their largest challenge — latency. Researchers from UC Berkeley and UC San Francisco have developed a new streaming approach using AI models that allows near real-time speech, a massive breakthrough from the previous 8-second delay for a single sentence. Read ➡️

Brain-Computer Interfaces for speech just overcame their largest challenge — latency. Researchers from UC Berkeley and UC San Francisco have developed a new streaming approach using AI models that allows near real-time speech, a massive breakthrough from the previous 8-second delay for a single sentence. Read ➡️

18,952 views

✨Just released: NVIDIA Nemotron post-training multi-lingual dataset ✨ We expanded permissive post-training dataset with addition of synthetically translated reasoning traces. ✋ Five new languages 💪 World class reasoning traces 🤗 Learn more: 📥 Download:

✨Just released: NVIDIA Nemotron post-training multi-lingual dataset ✨ We expanded permissive post-training dataset with addition of synthetically translated reasoning traces. ✋ Five new languages 💪 World class reasoning traces 🤗 Learn more: 📥 Download:

13,800 views

Hello from #GTC25 🤖 Send us your best jokes for the robot below 👇

Hello from #GTC25 🤖 Send us your best jokes for the robot below 👇

17,669 views

Excited to share our ICRA’23 IEEE ICRA work by Adithya Murali We scale up neural collision detection for object rearrangement with procedurally generated synthetic data. Project: Video: 🧵👇

Excited to share our ICRA’23 IEEE ICRA work by Adithya Murali We scale up neural collision detection for object rearrangement with procedurally generated synthetic data. Project: Video: 🧵👇

26,238 views

🎉 Congratulations Google DeepMind on the launch of Gemma 3 🎊 Introducing lightweight, multimodal, multilingual models streamlined for optimal performance on all your NVIDIA hardware – from your #datacenter GPUs to Jetson and RTX AI PCs for Windows. ➡️

🎉 Congratulations Google DeepMind on the launch of Gemma 3 🎊 Introducing lightweight, multimodal, multilingual models streamlined for optimal performance on all your NVIDIA hardware – from your #datacenter GPUs to Jetson and RTX AI PCs for Windows. ➡️

13,828 views

Developers at groundbreaking start-ups like Alpha3D and BOOM Interactive are making it possible to quickly and accurately convert 2D images into high-quality #3D models with #OpenUSD. Read more:

Developers at groundbreaking start-ups like Alpha3D and BOOM Interactive are making it possible to quickly and accurately convert 2D images into high-quality #3D models with #OpenUSD. Read more:

16,879 views

Videos

LIVE
1.2k

Anya Rossi

sweetdream.ai

SweetDream.aiSponsored•Livecam
Streaming Now

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

HD live stream
Exclusive private shows
1.2k viewers online

Free preview available • Premium content

🦞 Ready to deploy OpenClaw🦞? Our just released NVIDIA NemoClaw simplifies running OpenClaw always-on assistants more safely with a single command. ✅ Deploy claws more safely ✅ Run any coding agent ✅ Deploy anywhere Try with a free NVIDIA Brev Launchable: 🔗
2:33
NVIDIAAIDev's profile picture

🦞 Ready to deploy OpenClaw🦞? Our just released NVIDIA NemoClaw simplifies running OpenClaw always-on assistants more safely with a single command. ✅ Deploy claws more safely ✅ Run any coding agent ✅ Deploy anywhere Try with a free NVIDIA Brev Launchable: 🔗

NVIDIA AI Developer

189,223 views • 2 months ago

Our new open-source AI model family, NVIDIA Ising, will transform quantum computing by using agentic AI to solve it's most complex challenges, like real-time error correction. Nic Harrigan joins the NVIDIA AI Podcast to explain how this new open-source model family will help turn today's quantum hardware into tommorow's quantum-GPU supercomputers. 📺 Watch the full episode:
0:44
NVIDIAAIDev's profile picture

Our new open-source AI model family, NVIDIA Ising, will transform quantum computing by using agentic AI to solve it's most complex challenges, like real-time error correction. Nic Harrigan joins the NVIDIA AI Podcast to explain how this new open-source model family will help turn today's quantum hardware into tommorow's quantum-GPU supercomputers. 📺 Watch the full episode:

NVIDIA AI Developer

123,967 views • 1 month ago

Introducing NVIDIA Nemotron 3 Super 🎉 Open 120B-parameter (12B active) hybrid Mamba-Transformer MoE model Native 1M-token context Built for compute-efficient, high-accuracy multi-agent applications Plus, fully open weights, datasets and recipes for easy customization and deployment. 🧵
1:33
NVIDIAAIDev's profile picture

Introducing NVIDIA Nemotron 3 Super 🎉 Open 120B-parameter (12B active) hybrid Mamba-Transformer MoE model Native 1M-token context Built for compute-efficient, high-accuracy multi-agent applications Plus, fully open weights, datasets and recipes for easy customization and deployment. 🧵

NVIDIA AI Developer

134,803 views • 2 months ago

Securely experiment and build your own specialized agents, as the 671-billion-parameter DeepSeek-R1 model is now available as an NVIDIA NIM microservice in preview on Learn more ➡️
1:03
NVIDIAAIDev's profile picture

Securely experiment and build your own specialized agents, as the 671-billion-parameter DeepSeek-R1 model is now available as an NVIDIA NIM microservice in preview on Learn more ➡️

NVIDIA AI Developer

643,209 views • 1 year ago

The largest advancement of the CUDA platform since its creation in 2006 is here 👀 Introducing CUDA Tile, a tile-based programming model that provides the ability to write algorithms at a higher level and abstract away the details of specialized hardware, such as tensor cores. Read the technical blog 👉
2:34
NVIDIAAIDev's profile picture

The largest advancement of the CUDA platform since its creation in 2006 is here 👀 Introducing CUDA Tile, a tile-based programming model that provides the ability to write algorithms at a higher level and abstract away the details of specialized hardware, such as tensor cores. Read the technical blog 👉

NVIDIA AI Developer

244,885 views • 6 months ago

🦞 Set up NVIDIA NeMoClaw with MiniMax M2.7 (from MiniMax (official)) using secure, out-of-the-box integrations. ⚡Configure your environment and run search queries in minutes. 🎥 Full tutorial available here 👇 Or on YouTube 👉
5:11
NVIDIAAIDev's profile picture

🦞 Set up NVIDIA NeMoClaw with MiniMax M2.7 (from MiniMax (official)) using secure, out-of-the-box integrations. ⚡Configure your environment and run search queries in minutes. 🎥 Full tutorial available here 👇 Or on YouTube 👉

NVIDIA AI Developer

56,088 views • 1 month ago

Green and red have never looked so good. 💚🦞 NVIDIA CEO Jensen Huang surprises our panel guests at the #NVIDIAGTC keynote pregame. Watch the replay to hear the full conversation with Peter Steinberger 🦞, creator of OpenClaw🦞, on their partnership and the future of intelligent AI model. 🔗
0:27
NVIDIAAIDev's profile picture

Green and red have never looked so good. 💚🦞 NVIDIA CEO Jensen Huang surprises our panel guests at the #NVIDIAGTC keynote pregame. Watch the replay to hear the full conversation with Peter Steinberger 🦞, creator of OpenClaw🦞, on their partnership and the future of intelligent AI model. 🔗

NVIDIA AI Developer

56,708 views • 1 month ago

Just launched #CES2026, the new open-source NVIDIA Nemotron Speech ASR model is here to solve latency drift and redundant compute. Its cache-aware streaming architecture eliminates the need for buffered inference, giving you stable, sub-100ms latency (24ms median T-T-F) and up to 3x more throughput on your GPU. 🤗 Read the technical blog with real-world results from Daily and Modal on Hugging Face:
1:24
NVIDIAAIDev's profile picture

Just launched #CES2026, the new open-source NVIDIA Nemotron Speech ASR model is here to solve latency drift and redundant compute. Its cache-aware streaming architecture eliminates the need for buffered inference, giving you stable, sub-100ms latency (24ms median T-T-F) and up to 3x more throughput on your GPU. 🤗 Read the technical blog with real-world results from Daily and Modal on Hugging Face:

NVIDIA AI Developer

138,136 views • 4 months ago

Kimi K2.5 is now live on GPU-accelerated endpoints for free prototyping, so you can quickly start building with a frontier-scale multimodal model instead of just reading about it. Want to get started? We’ve got you: • Step-by-step tutorial • Ready-to-run GitHub notebook • First inference in minutes, not hours 📖 All available in the technical blog →
0:33
NVIDIAAIDev's profile picture

Kimi K2.5 is now live on GPU-accelerated endpoints for free prototyping, so you can quickly start building with a frontier-scale multimodal model instead of just reading about it. Want to get started? We’ve got you: • Step-by-step tutorial • Ready-to-run GitHub notebook • First inference in minutes, not hours 📖 All available in the technical blog →

NVIDIA AI Developer

108,397 views • 3 months ago

Here's your weekend project. Build a fully local, sandboxed AI assistant. Step-by-step tutorial to build your always-on agent: 🦞 on OpenClaw🦞 ✅ with NVIDIA NemoClaw ✨ using NVIDIA DGX Spark Get started:
13:49
NVIDIAAIDev's profile picture

Here's your weekend project. Build a fully local, sandboxed AI assistant. Step-by-step tutorial to build your always-on agent: 🦞 on OpenClaw🦞 ✅ with NVIDIA NemoClaw ✨ using NVIDIA DGX Spark Get started:

NVIDIA AI Developer

40,600 views • 1 month ago

🤯 100% of NVIDIA engineers code with AI—and they’re checking in 3x more code than before. At that scale, human-only code review can’t keep up. CodeRabbit review agents use models like Claude and GPT with NVIDIA Nemotron to: ✅ Pull context from code, docs, project trackers, and more ✅ Use Nemotron’s long context and reasoning to summarize ✅ …so that frontier models can flag issues and suggest fixes in minutes, not days​. Join us today at 11 AM PT to see an end-to-end demo and bring your Qs: 🤗 Get Nemotron Nano 3:
1:22
NVIDIAAIDev's profile picture

🤯 100% of NVIDIA engineers code with AI—and they’re checking in 3x more code than before. At that scale, human-only code review can’t keep up. CodeRabbit review agents use models like Claude and GPT with NVIDIA Nemotron to: ✅ Pull context from code, docs, project trackers, and more ✅ Use Nemotron’s long context and reasoning to summarize ✅ …so that frontier models can flag issues and suggest fixes in minutes, not days​. Join us today at 11 AM PT to see an end-to-end demo and bring your Qs: 🤗 Get Nemotron Nano 3:

NVIDIA AI Developer

100,151 views • 4 months ago

🧵 We are expanding our open model families—including Nemotron, Cosmos, and BioNeMo—to advance development in agentic AI, physical AI, healthcare, and more. 👇
1:08
NVIDIAAIDev's profile picture

🧵 We are expanding our open model families—including Nemotron, Cosmos, and BioNeMo—to advance development in agentic AI, physical AI, healthcare, and more. 👇

NVIDIA AI Developer

46,357 views • 2 months ago

✨Introducing Neuralangelo✨ A new #AI model to turn 2D video from any device -- cell phone to drone capture -- into 3D structures with intricate details. New from #NVIDIAResearch #CVPR2023 Read more:
0:57
NVIDIAAIDev's profile picture

✨Introducing Neuralangelo✨ A new #AI model to turn 2D video from any device -- cell phone to drone capture -- into 3D structures with intricate details. New from #NVIDIAResearch #CVPR2023 Read more:

NVIDIA AI Developer

497,873 views • 3 years ago

“Researchers need open source. Developers need open source. Companies around the world — we need open source.” -Jensen Huang #NVIDIAGTC
0:55
NVIDIAAIDev's profile picture

“Researchers need open source. Developers need open source. Companies around the world — we need open source.” -Jensen Huang #NVIDIAGTC

NVIDIA AI Developer

118,449 views • 7 months ago

🦞 Thanks to everyone who stopped by at Build-a-Claw yesterday. You can BYOD or purchase a DGX Spark there and engineers will help you install and answer any questions about OpenClaw🦞. 📆 Reminder - open each day of #NVIDIAGTC at 8am, until 5 p.m. Tues/Wed and until 2 p.m. Thurs. 📍 GTC Park 🔗
0:24
NVIDIAAIDev's profile picture

🦞 Thanks to everyone who stopped by at Build-a-Claw yesterday. You can BYOD or purchase a DGX Spark there and engineers will help you install and answer any questions about OpenClaw🦞. 📆 Reminder - open each day of #NVIDIAGTC at 8am, until 5 p.m. Tues/Wed and until 2 p.m. Thurs. 📍 GTC Park 🔗

NVIDIA AI Developer

45,623 views • 2 months ago

✨ Meet our new open family of models: NVIDIA Nemotron 3 Open in weights, data, tools, and training, Nemotron 3 is built for multi-agent apps and features: • An efficient hybrid Mamba‑Transformer MoE architecture • 1M token context for long-term memory and improved reasoning • Multi‑environment reinforcement learning via NeMo Gym for advanced skill adaptation Plus NVFP4 pre-training, latent MoE, 1T tokens of data, and more. 📗Read the details in our tech blog: 🤗 Try the model on Hugging Face:
1:35
NVIDIAAIDev's profile picture

✨ Meet our new open family of models: NVIDIA Nemotron 3 Open in weights, data, tools, and training, Nemotron 3 is built for multi-agent apps and features: • An efficient hybrid Mamba‑Transformer MoE architecture • 1M token context for long-term memory and improved reasoning • Multi‑environment reinforcement learning via NeMo Gym for advanced skill adaptation Plus NVFP4 pre-training, latent MoE, 1T tokens of data, and more. 📗Read the details in our tech blog: 🤗 Try the model on Hugging Face:

NVIDIA AI Developer

68,500 views • 5 months ago

Creativity unleashed 👨‍🎨🎨✨ 👀 Edify3D from #NVIDIAResearch explores a new way to generate text-to-#3D images. 📗 Project page: 📝 Paper: 🔊 sound on
0:42
NVIDIAAIDev's profile picture

Creativity unleashed 👨‍🎨🎨✨ 👀 Edify3D from #NVIDIAResearch explores a new way to generate text-to-#3D images. 📗 Project page: 📝 Paper: 🔊 sound on

NVIDIA AI Developer

177,483 views • 1 year ago

We’re excited to share that 🥇Llama Nemotron Super 49B v1.5 -- our latest open reasoning model -- is now #1 on the Artificial Analysis Intelligence Index - a leaderboard that spans advanced math, science, and agentic tasks, in the 70B open model category. Llama Nemotron Super 49B v1.5 is trained with high-quality reasoning synthetic data generated from models like Qwen3-235B and DeepSeek R1. It delivers state-of-the-art accuracy and throughput, running on a single H100. Key features: 🎯 Leading accuracy on multi-step reasoning, math, coding, and function-calling 🏗️ Post-trained using RPO, DPO, and RLVR across 26M+ synthetic examples 📊 Fully transparent training data and techniques If you're building AI agents and want a high accuracy, fully-open, and transparent reasoning model that you can deploy anywhere, try Super v1.5 on or download from Hugging Face 🤗 ➡️ Leaderboard:
0:54
NVIDIAAIDev's profile picture

We’re excited to share that 🥇Llama Nemotron Super 49B v1.5 -- our latest open reasoning model -- is now #1 on the Artificial Analysis Intelligence Index - a leaderboard that spans advanced math, science, and agentic tasks, in the 70B open model category. Llama Nemotron Super 49B v1.5 is trained with high-quality reasoning synthetic data generated from models like Qwen3-235B and DeepSeek R1. It delivers state-of-the-art accuracy and throughput, running on a single H100. Key features: 🎯 Leading accuracy on multi-step reasoning, math, coding, and function-calling 🏗️ Post-trained using RPO, DPO, and RLVR across 26M+ synthetic examples 📊 Fully transparent training data and techniques If you're building AI agents and want a high accuracy, fully-open, and transparent reasoning model that you can deploy anywhere, try Super v1.5 on or download from Hugging Face 🤗 ➡️ Leaderboard:

NVIDIA AI Developer

100,506 views • 10 months ago

We just released 3 million samples of high quality vision language model training dataset for use cases such as: 📄 optical character recognition (OCR) 📊 visual question answering (VQA) 📝 captioning 🤗 Learn more: 📥 Download:
0:29
NVIDIAAIDev's profile picture

We just released 3 million samples of high quality vision language model training dataset for use cases such as: 📄 optical character recognition (OCR) 📊 visual question answering (VQA) 📝 captioning 🤗 Learn more: 📥 Download:

NVIDIA AI Developer

95,389 views • 9 months ago

NVIDIA Cosmos Reason 2 is here. 🥳 An open, highly accurate reasoning vision language model for physical AI, featuring: ✅ Improved spatio-temporal understanding and timestamp precision ✅ Flexible deployment with 2B and 8B model sizes ✅ Long-context reasoning with up to 256K tokens ✅ Expanded visual perception across complex environments We also have new Cosmos releases: Predict 2.5, Transfer 2.5, and the NVIDIA GR00T N1.6 robot foundation model. 📗Read our technical blog: 🤗 Download Cosmos Reason 2 on Hugging Face:
0:57
NVIDIAAIDev's profile picture

NVIDIA Cosmos Reason 2 is here. 🥳 An open, highly accurate reasoning vision language model for physical AI, featuring: ✅ Improved spatio-temporal understanding and timestamp precision ✅ Flexible deployment with 2B and 8B model sizes ✅ Long-context reasoning with up to 256K tokens ✅ Expanded visual perception across complex environments We also have new Cosmos releases: Predict 2.5, Transfer 2.5, and the NVIDIA GR00T N1.6 robot foundation model. 📗Read our technical blog: 🤗 Download Cosmos Reason 2 on Hugging Face:

NVIDIA AI Developer

45,677 views • 4 months ago