Embedl

Team

company

https://www.embedl.com

embedl

Activity Feed

AI & ML interests

None defined yet.

Recent Activity

HProg updated a model about 8 hours ago

embedl/Cosmos-Reason2-2B-W4A16-Edge2

WilhelmT new activity 4 days ago

embedl/Qwen3-0.6B-FlashHead:about speed

WilhelmT updated a model 18 days ago

embedl/Cosmos-Reason2-2B-W4A16

View all activity

embedl 's collections 6

Cosmos-Reason2

nvidia/Cosmos-Reason2 multi-modal reasoning models optimized by Embedl.

embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
embedl/Cosmos-Reason2-2B-NVFP4A16

Image-Text-to-Text • 2B • Updated Mar 2 • 73 • 1
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12

NVIDIA Jetson AGX Orin

Models optimized and bench-marked for NVIDIA Jetson AGX Orin. Memory-efficient and latency-optimized variants designed for real-time edge inference.

embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
Running

4

Edge Inference Benchmarks

🚀

4

On-Device benchmarks across devices and models.

FlashHead

Efficient Drop-In Replacement for the Classification Head in Language Model Inference.

embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
embedl/Qwen3-1.7B-FlashHead-W4A16

0.8B • Updated Dec 16, 2025 • 3 • 3
embedl/gemma-3-270m-it-FlashHead

Updated Dec 20, 2025 • 8 • 3
embedl/Qwen3-0.6B-FlashHead

0.6B • Updated Dec 18, 2025 • 7 • 4

NVIDIA Jetson Orin Nano

Ultra-efficient model variants optimized for Jetson Orin Nano. Designed for constrained edge environments requiring low memory footprint.

embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
Running

4

Edge Inference Benchmarks

🚀

4

On-Device benchmarks across devices and models.

NVIDIA Jetson AGX Thor

Models validated and performance-optimized for NVIDIA Jetson AGX Thor. Tailored for high-performance edge AI workloads.

embedl/Cosmos-Reason2-2B-NVFP4A16

Image-Text-to-Text • 2B • Updated Mar 2 • 73 • 1
embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7

EdgeN

Quantization strategy where most weights are converted to INT4, activations remain in FP16, and sensitive layers are preserved in FP16.

embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7

Cosmos-Reason2

nvidia/Cosmos-Reason2 multi-modal reasoning models optimized by Embedl.

embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
embedl/Cosmos-Reason2-2B-NVFP4A16

Image-Text-to-Text • 2B • Updated Mar 2 • 73 • 1
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12

NVIDIA Jetson Orin Nano

Ultra-efficient model variants optimized for Jetson Orin Nano. Designed for constrained edge environments requiring low memory footprint.

embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
Running

4

Edge Inference Benchmarks

🚀

4

On-Device benchmarks across devices and models.

NVIDIA Jetson AGX Orin

Models optimized and bench-marked for NVIDIA Jetson AGX Orin. Memory-efficient and latency-optimized variants designed for real-time edge inference.

embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
Running

4

Edge Inference Benchmarks

🚀

4

On-Device benchmarks across devices and models.

NVIDIA Jetson AGX Thor

Models validated and performance-optimized for NVIDIA Jetson AGX Thor. Tailored for high-performance edge AI workloads.

embedl/Cosmos-Reason2-2B-NVFP4A16

Image-Text-to-Text • 2B • Updated Mar 2 • 73 • 1
embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16

Image-Text-to-Text • 2B • Updated 18 days ago • 917 • 6
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7

FlashHead

Efficient Drop-In Replacement for the Classification Head in Language Model Inference.

embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7
embedl/Qwen3-1.7B-FlashHead-W4A16

0.8B • Updated Dec 16, 2025 • 3 • 3
embedl/gemma-3-270m-it-FlashHead

Updated Dec 20, 2025 • 8 • 3
embedl/Qwen3-0.6B-FlashHead

0.6B • Updated Dec 18, 2025 • 7 • 4

EdgeN

Quantization strategy where most weights are converted to INT4, activations remain in FP16, and sensitive layers are preserved in FP16.

embedl/Cosmos-Reason2-2B-W4A16-Edge2

Image-Text-to-Text • 2B • Updated about 8 hours ago • 13.2k • 12
embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead

Image-Text-to-Text • 2B • Updated 18 days ago • 773 • 7

AI & ML interests

Recent Activity

Team members 6

embedl 's collections 6

Edge Inference Benchmarks

Edge Inference Benchmarks

Edge Inference Benchmarks

Edge Inference Benchmarks