First result using Qwen 3.8 Flash Next.
Using the endpoint that huggingface provided. I am currently trying a similar prompt both on my Mac as well as on my 5090 using the 1bit released quant. Results soon if it works.
Qwen 3.8 just went FREE on 2 platforms 😳
You can now test: - Qwen3.8 27B - DeepSeek V4 Flash - MiMo V2.5 - Qwen3.8 Flash Next FP8 TokenHarbor: → → base URL: Empero: → → API key: free → model:
Qwen 3.8 Flash Next is looking so goood!
I was wondering if 27b would outperform it but honestly they're so close it's like splitting hairs on design. Look at these beautiful websites!!
First Voxel pagoda test with Qwen 3.8-Flash-Next in OMP harness.
You have reached the end of the archive
All of qwen38