Can you build a robot AI agent to pick up your socks? 🧦
With MARS, Innate’s open-source robot, you can!
Thanks to @innate_bot for bringing MARS to the RTX Spark launch! A large model running locally on an RTX Spark laptop controlled the robot, from chatting with visitors to tidying up.
Show more
Complex transcription just got easier with NVIDIA Nemotron 3.5 Transcribe. 🎙️
Transcribe, translate, and identify speakers locally on RTX Spark and DGX Spark systems.
Meet NVIDIA Nemotron 3.5 Transcribe 🎙️
An open-weight, promptable 30B MoE model with ~4B active parameters that brings transcription in 40 languages, speaker attribution, translation, contextual prompting and audio understanding into one checkpoint.
Watch it keep track of who said what when everyone’s talking at once.
1/ 🧵
Show more
Neural rendering with NVIDIA DLSS and RTX Kit. On-device AI characters with NVIDIA ACE.
Six updates for game developers, including RTX Spark support, @UnrealEngine tools, and the latest open-source research on character motion.🧵
Show more
Built small. Built to specialize. ⚡
Congrats to @liquidai on LFM3! Compact, efficient models for text, vision, audio and more. Run them locally on your favorite NVIDIA hardware.
Introducing 𝙻𝙵𝙼𝟹: our new generation of Liquid Foundation Models. With a new tri-hybrid architecture and new multimodal encoders, the LFM3 family is the ideal model suite for specialized intelligence.
> 6 model instances, across text, audio, vision, and omni modalities, generative and decision models.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱: text, 256k context window.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱: vision-language model with video-input support for spatial understanding
> 𝙻𝙵𝙼𝟹-𝙰𝚞𝚍𝚒𝚘-𝟸𝙱: tool calls, voice chat, reasoning in 18 languages.
> 𝙻𝙵𝙼𝟹-𝙾𝚖𝚗𝚒-𝟸𝙱: an omni model combines vision, audio, and language.
> 𝚍𝟷-𝟸𝙱 and 𝚍𝟷-𝙾𝚖𝚗𝚒-𝟸𝙱: decision models with typed outputs and calibrated probabilities in a single pass. Omni adds audio input.
> fine-tune on 1 GPU.
> a new tri-hybrid architecture
> architecture optimized for 14 processor classes across GPUs, CPUs, and NPUs.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱 is a beast. It performs competitively with Gemma 4 12B (6x larger), and Gemini 3.5 Flash Lite, and GPT-5.6 Luna, with 1.9B parameters.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱 is #1# on the Pareto frontier of device intelligence 𝚙𝚒𝚙𝚎𝚝𝚝𝚎.𝚕𝚒𝚚𝚞𝚒𝚍.𝚊𝚒
Every modality for specialized intelligence. Pick the one that fits your use case, customize it, and run it on-device.
1/ 🧵
Show more
Stack Small. Scale Smart.
The ASUS Ascent GX10 64GB is built to scale with your AI workloads:
✅Flexible 64GB configuration
✅Stack 2 Ascent GX10 for 1.7× performance of a single 128GB system
✅Optimized for local AI and inference
Explore:
Show more
I was the opening act for a family of ducks.
Thank you, Berlin. What a GTC.
Europe is building its AI ecosystem and economy, from AI factories to agents to robots.
DSX POD: Vera Rubin modular factories built, cabled and tested in the factory, then delivered ready to produce intelligence.
Agents: models, harnesses, sandboxes, knowledge and tools. NVIDIA Agent Toolkit brings the pieces together so companies control what their agents know, access and do. NVIDIA OpenShell enhances safety.
Physical AI: new customers across AI-RAN, robotaxis and robotics.
Personal AI: after four years building with Microsoft, RTX Spark brings personal agents to the Windows PC.
Thank you to President @EmmanuelMacron, Prime Minister @andyburnham and Chancellor @bundeskanzler for hosting me. Great to celebrate Sferical One with Sweden’s technology leaders. And thank you to President @TokayevKZ for the warm welcome.
And to our partners and NVIDIA teams: thank you for making it happen.
The keynote is here:
Show more
Getting started with @UnslothAI on RTX Spark? Just run the .exe.
@danielhanchen and @NaderLikeLadder got hands-on with NVIDIA RTX Spark to talk about running and training AI models locally, exploring reinforcement learning, and making it easier to start building with Unsloth.
Show more
🦆 Put AI agents to work to help your robot learn to navigate.
Using Microduck as the reference robot, this walkthrough shows how to create a simulated obstacle course with @OpenAI GPT-6 Astra, build 3D assets with @nvidiaomniverse libraries, simulate physics with Newton, train policies in NVIDIA Isaac Lab, and validate them in Isaac Sim on @NVIDIARTXSpark.
Show more
RTX Remix brings path tracing and DLSS to Minecraft: Java Edition on RTX Spark
DGX Station for Windows just hit the stage with Jensen at GTC Berlin.
Coming later this year, build and deploy AI agents in the @Windows apps you already work in, powered by up to 748GB of coherent memory for trillion-parameter models at your desk.
Show more
DGX Station is bringing local AI to more teams and workflows.
Serve 30+ users, simplify IT management with new integrations, and build with @Windows native applications on upcoming systems expected later this year.
Show more
When @jensenhuang asks for a somersault, you deliver. 🦆
See how AI agents on NVIDIA RTX Spark help bring @huggingface's Microduck from simulation to the real world.
NVIDIA RTX Spark is here.
Build AI locally, bring your creative ideas to life, and take your favorite games along.
A new beginning for @Windows PCs starts today.
@NVIDIARTXSpark @ASUS Between referencing NVIDIA Nemotron models for my university ML Python projects and juggling DaVinci Resolve, Audacity, and Photoshop to produce my podcast episodes, this signed @NVIDIARTXSpark laptop would be the absolute ultimate daily driver! #ShareYourSpark#
Show more
Come hang out with me to celebrate the @NVIDIARTXSpark launch with demos, a meet and greet, and FIVE laptops signed by Jensen Huang up for grabs!
Friday, October 16, 10a at Best Buy, San Francisco! Click here for more! #ad#
Show more
Almost time! Windows laptops powered by @NVIDIARTXSpark arrive tomorrow.
Today we release 𝙻𝙵𝙼𝟸.𝟻-𝟸𝟺𝙱-𝙰𝟸𝙱, the largest 𝙻𝙵𝙼𝟸.𝟻 family member. It is an efficient local model for laptops and PCs to work through multi-step tasks inside popular agent harnesses.
> 24B total parameters, 2B active per token
> Context expanded to 131K tokens
> A 128K-token vocabulary
> 33T pretraining tokens
> It is a reasoning model
> Very easy to fine-tune and customize
> Performs competitively with Qwen3.6-35B-A3B, a model with ~50% more total and active parameters, on agentic and tool-use tasks.
> Open base and post-trained weights on @huggingface 🤗
1/
Show more
Congrats to @bfl_ai on the launch of FLUX 3 Image Klein! 🎉
More ways to create locally, with powerful image generation and editing in a model you can customize.
FLUX 3 Image Klein is here.
Open Weights with the quality of a large frontier model. Customizable and fast.
5.6B parameters and runs on a single consumer GPU.
Native 4K, bounding boxes, and pixel-perfect editing.
Show more
#ShareYourSpark#
I'm building Fern. This is a very unique desktop app that combines 1-click install for popular open-source image/audio/LLM models with a threads-focused UI/UX design. Imagine the Codex experience for local ai with zero setup required. I am currently building Fern on a 4 year old Lenovo X1 Carbon with an Intel Iris GPU and 32gb of ram.
Upgrading to the @ASUS ProArt P14 RTX Spark will allow me to fully build out Fern with maximal performance and integration flexibility. Fern will remove barriers to entry for local AI and make working with local AI models significantly easier for non-technical people.
Show more
NVIDIA Nemotron 3.5 Bolt 24GB puts powerful local agents within reach. ⚡
74B-A9B parameters packed into 24GB of memory, balancing intelligence and speed while running alongside your apps. Optimized for RTX Spark and ready for your next project.
Show more
