One-shot prompt test: Gemini 3.7 Flash (High Thinking) vs. Zcode + GLM-5.3
Glm-5.3: 10 mins, burned 430k tokens. looks basic but the animation is actually solid • gemini 3.7: 3 mins in ai studio. way more control as usual, but the sim feels kinda arcade-y tbh Is it
Qwen 3.8 max + deepseek V4 pro for free
Found another way to try some powerful AI models without paying for another subscription. How to get it: Go to Log in to your account Open the models section Pick a model and start using it Right now, Qwen 3.8
Made by Qwen 3.8 27B 8-bit on Mac Studio M3 Ultra 512G.
I am impressed that I can accomplish this with a local model - excited to explore how far I can push it. Alibaba_Qwen Prompt: make me a simple budgeting app.
Atelier MIRA is now compatible with Qwen 3.8 27B.
Image recognition and prompt creation can now be completed with just Atelier MIRA! This video was also made with Atelier MIRA!
You have reached the end of the archive
All of qwen38