Introducing FastH3-Omni.
The fastest and highest-quality open-source version of realtime H3 by @MiniMax_AI, post-trained by FastVideo @haoailab with @nuvalab.
Exclusive preview on the Reactor API until 10.20.2026.
• First and last frame conditioning
• Audio input + strong dialogue
• Up to 3 reference images
• Frontier-level performance
introducing muse wonder
our new frontier model. state-of-the-art at personal agents and professional agents.
our largest and most capable model yet—rivals the big bois
Berlin, you look good in CUDA. 💚
#NVIDIAGTC is coming to town. We thought we’d say hello from the TV Tower.
Spot it? Tag @NVIDIA and show us your view. #FestivalOfLights
Historically, to create an animation like this, you always need a set of tools:
After Effects, Blender, Cinema 4D.....
Today, you can do all of that just using Claude and Higgsfield Katana.
No video generations. Pure code.
Nano Banana 2.1 is available on VidGuy!
- More control over complex prompts
- Faster generation
- Higher quality across the board
- Character consistency from one edit to the next
Generate in 4K and post straight to TikTok on autopilot.
Only with VidGuy
Meet Mistral Large 4, aka Le Chonk.
• 1T parameters, natively multimodal. 49B active.
It is the best open weights model from US or Europe on aggregated benchmarks.
• State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding.
• Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure.
• Available to all via API today. Working with cybersecurity partners privately.
Open weights release end of October.
1/ Agent cursors were never much to look at.
Today we're introducing 6 new cursor motions in Cua Driver, each hand-crafted and perfectly timed. We studied how people aim a mouse, built 82 motions in a playground, and kept the best six.
It's all open source: https://t.co/0R5fWjfXvO