autotrust/Qwen3.8-Flash-Next-E384-FP4-V2
image-text-to-text · ♥ 2.4K · ↓ 42.7K
- Taskimage-text-to-text
- Downloads42.7k
- Likes2370
- Licenseother
image-text-to-text · ♥ 2.4K · ↓ 42.7K
A compact build of nvidia/Qwen3.8-Flash-Next-FP4-Local found by Neural Architecture Search (NAS) over the routed-expert configuration: it keeps 384 of the 512 routed experts in every decoder MoE layer and in the MTP layer.
Unchanged from the original: - top-10 routing and the shared experts; - attention, the n-gram PLE table, the vision tower and the tokenizer.
Excerpt from the model card on Hugging Face ↗.
autotrust/Qwen3.8-Flash-Next-E384-FP4-V2 was first seen on Hugging Face · Trending models on Nov 9, 2026. Since then it has appeared in 56 checks on 3 different days. Its best position was #25 of 30, reached on Nov 10, 02:42 UTC. It dropped off the list after Nov 11, 16:42 UTC.
Position on Hugging Face · Trending models at every check (lower is better).
Best rank per day (UTC).
Platform metric as reported by Hugging Face.
Archived daily, weekly and monthly lists that include this entry.
Not at the moment. It was last on the list on Nov 11, 16:42 UTC, after 3 days of trending.
#25 on Hugging Face · Trending models, on Nov 10, 02:42 UTC. Trendoday checks this list every 60 minutes.
It was first seen on Nov 9, 2026 and has been on the list on 3 days (56 checks) since.
It is tagged image-text-to-text with vllm on Hugging Face and has 42.7k downloads under the other license.