Join us Tuesday in Brooklyn for an evening with AI leaders and practitioners. CEO William Falcon will sit down with NYU professor KyungHyun Cho for a fireside chat on building/deploying modern AI systems at scale + networking & drinks. Seats are going fast! Register now → go.lightning.ai/4a8oYHM

We’ll be at Cisco Live EU in Amsterdam showing teams how to train, deploy, and operate AI in production. If you’re attending Feb 9–13, visit our booth or lock in time to meet with the team → go.lightning.ai/4sQPmgL

Today, Lightning AI & Voltage Park have merged bringing together AI software & owned GPU infrastructure into a single, AI-native cloud, for training, deploying, & running AI at scale. Excited for what this unlocks for developers, enterprises, & the future of AI. Read more → go.lightning.ai/4sK6STH

Community spotlight: Lightning Pose 3D is live 🚀 Our friends Matt Whiteway and the Paninski Lab just shipped a complete redesign of the popular pose estimation tool, built for real-world, multi-camera data. This is what production-ready research tooling looks like when it's built on Lightning.

🎉 Excited to announce Lightning Pose 3D! A complete redesign of our app with powerful new features: calibration-aware labeling for multiview data, state-of-the-art multiview transformer models for training & inference, an interactive video viewer, and continued support for single-view projects.

Join us Thursday, May 15 in San Francisco for a hands-on workshop led by William Falcon, creator of PyTorch Lightning and CEO of Lightning AI. Learn how to create lightning-fast, scalable AI agents that are production-ready from day one. Space is limited—RSVP now: lu.ma/2enz2bx4

🚀 Deploy Qwen3 - the smartest, cheapest open model - on your cloud in hours. Try it now ➡️ go.lightning.ai/3SjVoFs ✅ MoE with 131K context and 22B active params ✅ Thinking optional = lower costs ✅ No lock-in, private, SOC2 & HIPAA compliant ✅ Use startup credits or cloud commits

In the most recent Lightning Thunder Session, Luca and Tom look at doing DDP on the sharded weights from FSDP in a 2d device mesh. Watch the full episode here ➡️ youtu.be/iIMA9_ippEw

Last month, we hosted a Robotics Meetups in New York City featuring talks from 2 industry experts: ✨ Lerrel Pinto, NYU Assistant Prof. of Computer Science ✨ Shagun Majotra, Mechatronics and Robotics Engineering Masters Student This clip explores constructivism's relationship to building robots:

🚀 Deploy @Alibaba_Qwen's Qwen3 - the smartest, cheapest open model - on your cloud in hours. Try it now ➡️ go.lightning.ai/3SjVoFs ✅ MoE with 131K context and 22B active params ✅ Thinking optional = lower costs ✅ No lock-in, private, SOC2 & HIPAA compliant ✅ Use startup credits or cloud commits

🚀 Deploy @Alibaba_Qwen's Qwen3—the smartest, most cost-efficient open LLM in hours. ✅ Long 131K context, and massive 22B active params ✅ Control how much it thinks to reduce cost ✅ Agent-ready & multi-lingual ✅ Secure, private, cost-efficient Production-ready in <24h ➡️ go.lightning.ai/3SjVoFs

Join Lightning AI’s open source contributor program to work on PyTorch Lightning, Fabric, LitGPT, LitServe, LitData, TorchMetrics and Thunder. Collaborate with top engineers and researchers to power AI used by 30K+ organizations.

Don't miss our CTO, Luca Antiga, speak next week at PyTorch Day in Paris on May 7 🇫🇷 Learn how Lightning Thunder makes it easy to write custom code transformations that scale models and cut GPU waste without sacrificing performance or usability. #PyTorchDayFrance

⚡ Persist your PyTorch Lightning checkpoints on Lightning AI 🧠 Pause and resume training anytime 📊 Track performance across runs 👥 Share with your team, org, or make it public Learn more ➡️ github.com/Lightning-A...

🚨 B200s are here — train & deploy 3x to 4.5× faster than H100s with Lightning AI and our compiler, Thunder. Optimized for: ✅ Real-time robotics and edge AI ✅ Science workloads ✅ Massive LLMs Reserve or run instantly, no lock-in → lightning.ai

We had a blast 🚀 engaging with the Robotics community at the New York City and London meetups this week! Featuring insightful presentations from NYU's @lerrelpinto.com and Shagon Majotra, along with food and drinks. Stay tuned for info on our next Robotics meetup in San Francisco next month.

Hey YC founders 👋 If you're building with AI, Lightning AI has some great perks for you: • 3 enterprise-tier seats for $10K (normally $30K) • $5K in compute credits • Showcase your startup through case studies or events Let's build faster together ⚡️ lightning.ai/yc-startups

Deploy state-of-the-art object detection w/ RF-DETR! Learn how to serve Roboflow’s RF-DETR model using LitServe for real-time object detection — no heavy lifting required. 🚀 Achieves 60+ average precision (AP) on the COCO benchmark 🎯 Processes 25 images per second on a T4 📦 Built for the edge

The PyTorch Day France 2025 schedule is now live: Explore the full agenda of talks and sessions ⛓️ pytorchdayfrance2025.sched.com Co-located with #GOSIMAI2025 🎟️ Use code PYTORCHFRIEND for 25% off registration 👉 Or enter the Lucky Draw: paris2025.gosim.org #PyTorch #PyTorchDayFrance

There’s still time to RSVP to tomorrow's Robotics meetups in London and NYC! London: lu.ma/LondonRobotics NYC: lu.ma/NYCRobotics

Lightning AI can cut deployment costs by 15% with fast and measurable cold-starts: ✅ Enterprise-grade reliability ✅ Scale to zero to save money ✅ Move between multiple clouds seamlessly Check it out 👉 lightning.ai/lightning-a...

Automate your entire Lightning workflow with the Lightning CLI — spin up Studios, run jobs, execute commands, and more. Get started in seconds: `pip install -U lightning-sdk` Log in to your Lightning account via the CLI: `lightning login` Get started:

Learn how to optimize costs with this Lightning Studio template featuring 3 hands-on notebooks using NVIDIA cuML: ⚡️ Lightning-fast classification 📊 Scalable clustering 🔍 Effortless dimensionality reduction Try it now 💥 go.lightning.ai/4cKTAPs

The Lightning platform now tracks CPU/GPU utilization across teams to optimize efficiency week after week. 📸 Get a snapshot of active Studios 📊 Track and optimize machine utilization ⏲️ Monitor and shorten cold start speeds Turn insights into savings ➡️ go.lightning.ai/3RqSXR8

We're just one week away from our monthly meetups in London, NYC, and SF! Join us to hear from top speakers & industry leaders in robotics as they share insights on what's next in the field. RSVP today & we'll see you there! ⚡️

🚨 OSS update: We've added Equal Error Rate (EER) to TorchMetrics. Now it’s even easier to get accurate, reliable evaluations on biometric systems or ROC analysis. Thanks to Nicki Skafte for the contribution! Check it out here: github.com/Lightning-A...

Secure & rate-limit LLM APIs with LitServe middleware! 🚀 Control API usage efficiently 🔐 Add security layers 🌟 Measure, log, and customize requests Easily extend deployment with just a few lines of code. Try it out 👉 github.com/Lightning-A...

Save money with same day reservations of in-demand GPUs. Instantly reserve A100, H100 and H200 machines for half the price!

This week's Thunder Session had @lantiga and Tom look forward to having a joint forward and backward trace to work with. Check out the full episode ➡️ youtu.be/NGKIDxzjNR0

In the last Thunder Session, Luca and Tom discussed state —model params + metadata needed to save/load during training - and how it gets transformed in the FSDP Transform. Thunder Session livestream returns Friday. Watch it live here on X or youtube: www.youtube.com/live/Hn9isu...

🚀 Instantly deploy Meta’s latest flagship reasoning model, Llama 4 API, for visual question answering, image captioning, and multimodal content generation. Easily integrate Scout and Maverick into your applications with Lightning AI Hub’s enterprise-grade API.

Curious about the Lightning AI stack? 🤔 Learn about the full life cycle, from data preparation to deployment, with CTO Luca Antiga at his presentation at #GTC25. Check out our GTC talk here: www.nvidia.com/en-us/on-de...

🚀 Train models efficiently with this Lightning Studio template created by a community member, Deependu Jha. From data to evaluation, it’s got you covered: ✅ LitData for data streaming & optimization ✅ PyTorch Lightning for structured training ✅ TorchMetrics for evaluation & metrics

🔥 LitServe just hit 3K GitHub stars! 🚀🎉 Why are developers loving LitServe? ✅ Blazing-fast AI model serving ⚡ ✅ Pythonic & easy to use 🐍 ✅ Deploy anywhere—cloud, on-prem, or edge 🌍 ✅ Backed by an amazing community 💜 Try it out 👉 github.com/lightning-a... #AI #MachineLearning #MLOps #LitServe

In case you missed it, Luca & Tom optimized PyTorch models using Thunder at #GTC25. ⚡⛈️ Learn how to transform & distribute models while keeping your code unchanged - check out the full talk here ➡️ www.nvidia.com/en-us/on-de...

It’s offsite week at Lightning AI! Jam packed a lot into Day 1. Team bonding, goal setting and strength training 💪

Simplify your Jira workflow. 🚀 Streamline complex projects, ticket navigation and boost productivity with this RAG-powered AI studio. No digging needed - just get fast, accurate, context-aware answers.

We’re going deep on Robotics this April. If you’re in NYC, the Bay Area, or London and want to feature your work, shoot us a DM—meetups in each hub are happening at the end of the month. ⚡

Lightning’s Studios now offers reserve-ahead for A100s, H100s, and H200s for 50-80% off! ✅ Reserve by day up to a month  ✅ Self-serve with pay-up-front billing ✅ Reservation prices up to 80% off Reserve yours ➡️ t.co/ES7E6GmLlq

Deploy a background removal model. 🚀 Created by community member Sitam Meur, this studio uses BEN2, a deep learning model for background removal and foreground extraction, utilizing Confidence Guided Matting (CGM) for precise cut-out effects. Check out this self-hosted/private API with LitServe.⚡

Segger, a graph neural network that redefines transcript-to-cell assignment in spatial transcriptomics, solves common issues of cell segmentation. Faster, more accurate, & scalable—Segger integrates seamlessly into workflows. And it’s trained on PyTorch Lightning ⚡

Deploy an API template with Microsoft's Phi-4 Multimodal. 🚀 At just 5.6B params, this compact model for speech, vision and text is designed to push the boundaries of multimodal research. Self-hosted, lightning-fast & OpenAI-compatible— deploy on Lightning AI Hub! ⚡

🚀 Build a corrective RAG (CRAG) agentic workflow, powered by DeepSeek-R1! Search your docs, evaluate context quality & fall back to web search when needed. Plus, a sleek @streamlit UI for seamless interaction. 🔥

🚀 Don't miss Luca Antiga, CTO of Lightning AI, in his #GTC25 speaking session: Train and Serve AI Systems Fast With the Lightning AI Open-Source Stack. ⚡️ 11:20 AM - 11:35 AM PDT ⚡️ San Jose Convention Center Grand Ballroom Theater (L2)

🚀 Join Luca Antiga & Thomas Viehmann for their talk, Make My PyTorch Model Fast, and Show Me How You Did It TODAY at #GTC25! Learn how Thunder - PyTorch-to-Python compiler - optimizes PyTorch models with ease without changing the code!

Join us for AI After Hours, hosted by Lightning AI at #GTC25! Connect with top AI experts, network with peers, and dive into the latest AI advancements in an exclusive, invite-only setting. ⚡️ Thursday, March 20th | 7:00 PM PT⚡️ Spaces are limited—secure your spot here go.lightning.ai/gtc-mixer-25

🚀 Excited to be at #GTC2025 this week! Stop by booth #2403 to meet our team and explore how to scale AI from research to production. Don’t miss our live demos and more! See you there! ⚡

Unlock the power of Fully Sharded Data Parallel (FSDP) & Thunder! ⚡ Optimizing bucketing logic boosts tensor materialization on faster machines—cutting overhead & scaling efficiency. Watch the full Thunder Session with Luca Antiga and Thomas Viehmann 🔗 in the comments.

🚀 LLaDA: Diffusion Models for LLMs challenges the dominance of autoregressive models in LLMs! LLaDA, a diffusion-based Transformer, achieves strong in-context learning, competitive instruction-following, and even beats GPT-4o in reversal tasks. Trained with PyTorch Lightning ⚡

Join us in person this week in #London, #NYC and #SF to learn how to deploy @deepseek.bsky.social securely on our cloud our yours. RSVP asap as spaces are filling up! London Tues Feb 4 go.lightning.ai/3Q04SVg NYC Thur Feb 6 go.lightning.ai/42GL0hE SF Fri Feb 7 go.lightning.ai/3Q4415S