Qwen 3.8 - 27B MLX-Serve 4bit model, made the best version of the famous Pagoda test so far for me.
Using pidotdev as the coding agent, and MLX-Serve as the backend.
Qwen 3.8 27b can generate A LOT more 0 shot games than Space invaders, Tetris, and Flappy Bird!
Alright guys, wow. Qwen 3.8 27B is basically Qwen 4.
What we have here is one of the best one-shot shark survival games I've made with any model. I ran this same prompt through fable, and the biggest thing all models struggle with in this test is the top-down view of the
Qwen 3.8 27B (dense) running on a single RTX 4090 (24GB VRAM) at 65 tokens/sec decode with MTP!
260,000 context window or 65 tokens/sec decode with native MTP. The API cartel should be terrified. We are officially running frontier tier agentic AI (benchmarks comparable to
You have reached the end of the archive
All of qwen38