Grok Imagine Image 2.0 is live.
Magic Wand for precise edits.
Sharp text.
Multi-ref (up to 5 images).
Smart Resize.
#2 in the world for generation and editing.
Built for actual work, not just pretty demos.
Quality Mode is live now on https://t.co/BkdWuqWhGT and the apps.
Had a lot of fun animating our new Grok Bot icon. It's entirely made in code, with every state, shape and expression able to morph smoothly into the other. And of course, I made it with Grok Build.
One of the coolest parts of Grok Bot is how customizable it is
You can create completely different Bots for different jobs, give each one its own identity and workflow, and then power them using your existing SuperGrok subscription
Grok Imagine Updates
- New 5:2 aspect ratio, great for banners, wide-format designs, and panoramic compositions.
- You can now generate only 2 images (previously, the minimum was 4).
Try it for free for a limited time.
Nemotron 3.5 Lightning fits real agent systems: not costly giant frontier model, but fast open 30B MoE (~3B active) for high-vol exec layer. Continuous agents mostly call tools, check results & loop, optimized for that, w/ up to 4x speed & 1M ctx legit for size.
Fully open (wts, data, recipes) under permissive lic & easy post-train matters. NVIDIA treats SW stack seriously, not just GPUs. Hybrid Mamba+MoE smart for long-run eff.
Won't dethrone top reasoning models on pure intel. But as specialized always-on workhorse from 1 high end GPU to DC? Useful. Eager for fine tunes in prod agent loops.