NVIDIA Nemotron 3 Ultra is now live on Lightning ⚡
• Open frontier reasoning for long-running autonomous agents
• Up to 1M-token context
• Up to 5x faster inference
• Up to 30% lower cost for agentic workloads
Deploy and orchestrate multi-model agent systems across cloud, hybrid, and on-prem environments.
Try it → https://t.co/bbvOezjQUF
Was a ton of fun participating in @nvidia Inference Day. Great chance to see how deeply the open source community is impacting the inference space. Some really great things to come from the @LightningAI optimization and platform teams! Keep your eyes here 👀
Yesterday we saw a supply-chain attack on PyTorch Lightning (on Pypi, not our core repo). It's wild how it happened but it was caught and quarantined within 42 minutes thanks to the open source community.
It's one reason why open source actually helps increase the security posture of projects. Thank you to @pypi and @SocketSecurity ⚡️
Summary: On April 30th, 2026, an attacker captured PyPI credentials and used them to push compromised versions of PyTorch Lightning (PTL).
These versions were live for 42 minutes before PTL community members alerted us and PyPI quarantined the package. The PyTorch Lightning GitHub source code repository was never compromised. This affected those who installed PTL via PyPI between 12:45:20 and 13:27:30 UTC on April 30th, 2026.
https://t.co/dcdhF6zwkh
Everyone wants to build multimodal agents but most get stuck stitching together separate vision, audio, and text models.
Run @nvidia Nemotron 3 Nano Omni on Lightning. See, hear, and reason in a single workflow.
One model instead of many. Higher throughput. Open weights, lightweight enough to run continuously.
Deploy now and get 30M free tokens monthly → https://t.co/MZMUYvxbE8
Anthropic just bought a NYC summer’s worth of data center power.
~10GW from Big Tech (AWS + GCP)
NYC peak summer: ~11GW
People are right, AI isn’t scaling like any software before it…it’s scaling like a city.
p.s. this should be the measurement of data center expansion / investment the way that “they’re 4 uber’s big” is how Robinhoof Snacks used to qualify companies.
Most models break as context grows.
DeepSeek AI V4 Pro doesn’t.
✅ 1M token context
✅ Mixture-of-Experts routing
✅ Stronger tool use across long workflows
Built for agents that don’t reset every few steps.
Now available for inference on Lightning. Day zero access.
Run it → https://t.co/u9tZCyR2r5
This morning is serving up all kinds of wins ⚡️
- First major provider to drop Deepseek support
- Top of the leaderboard for Nemotron 120b
- Baby slept 11 straight hours last night
🚀 Sneak peek of Deepseek v4 Pro on @LightningAI - clearly our optimization team had a post 5pm espresso so you could try this out before your 7am espresso. Link below:
i was recently introduced to someone with the coldest line i’ve heard in a long time.
“he’s building the last file system”
instant legend (both the person making and being intro’d)