Community
Reddit thread asks whether Qwen4 will clearly outperform Qwen3.8-Flash-Next
Community discussion speculates on Qwen4 size and quality relative to the current Qwen3.8-Flash-Next preview model, which is described as strong for its size. No official Qwen4 information is given.
Every outlet that ran it 1 article
r/LocalLLaMACommunityLead