Product 01
Yeti Claw
Editorial front door, chat command center, and public prompt surface for the NeonFlux fleet.
Yeti Claw + Yeti Cloud
Yeti Claw
The most savage AI agent.
A cold-blooded public inference surface for Spark image generation, Mac mini chat models, and fleet-routed prompts across web and apps.
Product 02
Yeti Cloud
Routing layer behind Spark, the Mac mini, and BeastMode inference lanes across every client surface.
Qwen Image on DGX Spark
Command Center
Prompt first, then read the archive when you need the deeper operator context.
The live chat surface sits here near the top on purpose. Pick the host, pick the model, and work inside one unified thread without the prompt box drifting away from the response stream.
Live Thread
Loading conversation surface...
The selected model reply stream appears here.
Sync: waiting for first inventory check
Yeti Claw CLI
Yeti Cloud Linked
Prompt The Beast
Type directly into Yeti Claw.
Route commands into Spark image models and the Mini fleet from one brutal command surface.
Yeti Claw
Get Oriented
Choose the right lane before you prompt.
Start with Mission Control if you want benchmarks and operator notes, switch to Gojira for Spark image work, use BeastMode for ESXi-backed inference lanes, or drop straight into the command center when you already know the model you want.
Stay Inside The Site
Read the findings, open the lane, then prompt the model in context.
Every major dispatch links directly into the surfaces it discusses, so operators can move from benchmarks to real prompts without losing context.