How long before we have a local model capable of modeling?

A user on reddit.com is discussing the current capabilities of local models for 3D modeling, noting that models like Qwen 3.8 and Qwen Flash, even on a powerful 5090 GPU, are significantly behind the performance of the latest Opus 5.5. They express that while a remotely capable local model would be very helpful, current local models are not yet close to meeting their needs for 3D modeling tasks.

This discussion highlights the current performance gap between local models like Qwen 3.8 and cloud-based models such as Opus 5.5, unlike earlier discussions that focused on general AI capabilities.