AI Daily Brief — 24 March 2026

NVIDIA expands the NemoClaw rollout with new optimisations for local agents on RTX PCs and DGX Spark. The supported open models for local agents now include Qwen 3.5 (27B, 9B and 4B variants) and Mistral Small 4.

Top stories

• NemoClaw + Qwen 3.5 / Mistral Small 4. NVIDIA validates new open models on its RTX PC and DGX Spark local-agent stack. via NVIDIA
• Qwen 3.5 VLM endpoints. Native multimodal agent development with Qwen 3.5 VLM is shown on NVIDIA GPU-accelerated endpoints. via NVIDIA Technical Blog

Compiled by AI Feed’s editor from publicly reported announcements on 24 March 2026.