We’re launching Gemma 4 12B: Our unified, encoder-free model that brings powerful multimodal intelligence straight to your laptop 🚀
The model bridges the gap between our mobile E4B model and larger 26B MoE models, packaging frontier-class reasoning and native audio into a highly optimized footprint, all under a permissive Apache 2.0 license.
Here’s what makes it unique:
+ Encoder-Less Architecture: We removed the multimodal encoders. The vision and audio inputs flow directly into the LLM backbone.
+ Agentic Performance (16GB VRAM): Run complex, multi-step workflows locally, with performance nearing our 26B model.
Reflective street signs are designed to be seen by drivers: engineered to bounce light back at you so they show up brightly, with minimal illumination. That makes them easy targets for a lidar's reflectivity channel. But reflectivity only tells you that something reflects. It doesn't tell you what it is, or what it means.
This is a Rev8 OS1 Max driving down Potrero Avenue in San Francisco. The same drive shown first in reflectivity, then in native color. The second the channel flips from reflectivity to RGB, traffic cones, signal phases, detailed signage, pedestrians, and road features all become visible as semantically meaningful objects.
$NVDA CEO Jensen Huang said:
“The next frontier of AI is physical AI. Every industrial company is becoming a robotics company.”
And $OUST is sitting right at one of the biggest bottlenecks of that future:
LiDAR sensors — the eyes of autonomous robots.
$OUST is the only durable Western LiDAR company left in the sector.
- Before the market understands: <$40
- After the market understands: $120
Your thoughts on $OUST?