Verified reference stack
Developer — Agentic Coding — RTX 4090
Current OpenHands guidance calls for at least 24 GB VRAM for quantized variants.
LM Studio local server
Qwen3.6-35B-A3B
RTX 4090 24 GB / 64 GB RAM
GOOD_QUANTIZEDLocal inference configuration
Apache-2.0
Evidence
Freshness rule: Material changes set this stack to REVALIDATION_REQUIRED.