Real world use of QWEN 3.8 local model on an MBPro M5 Max 128GB
I got overhyped by all the posts it’s slow… don’t expect it to build anything meaningful in 1 day unless you have some beast machine
I tested the same 3D generation prompt on 4 AIs
Gemini 3.7 Flash finished in less than 2 min. Qwen 3.8, Grok 4.6 and GLM 5.3 went over 6 min. But in terms of quality, Qwen 3.8 won again. Speed or finish: which do you choose?
Trying to make a movie locally on RTX 5090 32GB
Qwen 3.8 for scripting and screen play qwen image edit for images new minimax h3 for image to video let’s see how it goes and if i get the success stay tuned in so far seems promising
Can you spot the difference?
👀 Speculative decoding helps accelerate generative AI at the edge, especially on NVIDIA Jetson. Qwen 3.8 27B went from 13 to 35 tokens/sec, while Nemotron 3.5 Lightning went from 65 to 115. See how it works:
You have reached the end of the archive
All of qwen38