Ornith 35B/3B MoE Runs Fast and Accurate Locally

Developers tested the Ornith 35B-parameter MoE model with only 3B active parameters, running locally via ollama at about 23.5GB VRAM and reaching 78 tok/s on an NVIDIA RTX Spark N1X with perfect accuracy on simple tasks.

2026-10-19 ~ 2026-10-19 · 2 related posts

• Ornith 35B-A3B MoE hits 78 tok/s with 100% accuracy on NVIDIA RTX Spark in first tests — draginol · 2026-10-19
• 35B-param MoE model with 3B active runs local at crazy speed, 100% accurate on simple tasks — draginol · 2026-10-19