Ornith-1.5-35B-A3B-GGUF
★ agentic★ chat★ coding
Overview
Ornith-1.5 35B MoE (~3B active) at Q4_K_M - SWE-bench Verified 79 and GPQA Diamond 89.2 from a 22 GB download that fits a single 24 GB GPU.
Good for
- agentic
- chat
- coding
Recommended hardware
≈ 24 GB VRAM to run this 35B model at the Q4_K_M quantization.
Fits 24 GB+ consumer GPUs (e.g. RTX 3090, RTX 4090)