VIP

kennel

local.ai/kennel

Claimed August 2026

Local LLM inference and benchmarking, AI coding agents, GPU optimization on consumer and datacenter hardware, multimodal models, edge AI, and practical ML experimentation.

My machines

NVIDIA RTX PRO 6000 Blackwell 96GB
NVIDIA · 96 GB
Recommended for NVIDIA RTX PRO 6000 Blackwell 96GB
Choose what matters most
Best fit
Qwen3.6 35B A3B · NVFP4
Best balance of speed and intelligence.
Intelligence69.7
Speed47.7 tasks/hr
Size23.4 GB
More intelligent
Qwen3.6 27B · FP8
6% more intelligence · 86% slower.
Intelligence73.7
Speed25.6 tasks/hr
Size30.9 GB
Faster
Ornith 1.0 35B · Q4_K_M
36% faster · retains 97% intelligence.
Intelligence67.2
Speed75.0 tasks/hr
Size21.2 GB
Compare all models
P40
×2
GPU · 24 GB each
We don't have a benchmark for P40 yet.
RTX 4000 Ada
GPU · 20 GB
We don't have a benchmark for RTX 4000 Ada yet.
Ryzen Zen 4 7000 (Ryzen 7)
CPU · 64 GB
We don't have a benchmark for Ryzen Zen 4 7000 (Ryzen 7) yet.
Ryzen Zen 4 7000 (Ryzen 7)
CPU · 32 GB
We don't have a benchmark for Ryzen Zen 4 7000 (Ryzen 7) yet.

kennel

Bio

My machines

Choose what matters most