ModelFirst trended Aug 25
Daily trend score. Dots are days in the top 20; a ringed dot is a day at number one.
What carried it
- GitHubMiaAI-Lab/Qwen3.8-Flash-Next-Single-DGX-Spark: Qwen3.8-Flash-Next on ONE DGX Spark (TP=1)Sep 4 · 492 stars
- HNShow HN: Running 104GB Qwen3.8-Flash-Next on 48GB Mac with at ~12 tok/sSep 1 · 299 points and comments
- GitHubcarloslfu/slotstream: Run a 105 GB AI model on a 48 GB Mac. Qwen3.8-Flash-Next (125B mixture of experts) streams its experts from SSD through a slot cache. Native Swift on MLX and Metal: one binary, nAug 28 · 391 stars
- GitHubpeonist-ai/halogen-flash-server: The fastest way to run Qwen3.8-Flash-Next on Strix Halo (gfx1151)Aug 26 · 668 stars
- HNQwen3.8-Flash-NextAug 26 · 821 points and comments
- HNQwen3.8-Flash-Next Technical Report [pdf]Aug 26 · 36 points and comments
- HNQwen 3.8-Flash-Next releasing tomorrow (125B a6B)Aug 25 · 423 points and comments
Mac
DGX Spark
Python
Ollama
OpenAIDays in the top 20
Trend scores are computed from public engagement signals on the sources shown. Logos via logo.dev; trademarks belong to their owners.