I released an updated family of turboexl3 + mlx-serve streaming packs for DeepSeek 4.1…
This is a AI post classified by AGK Intelligence as Open & local models (a model release), kept by the AI Radar because it carries real work, not commentary.
I released an updated family of turboexl3 + mlx-serve streaming packs for DeepSeek 4.1 1017! These are designed for M5 96gb and 256gb devices wanting to run DS4.1 at reasonable speeds (70tps/1900tps) and reasonable quality (around NVFP4 equivalents) as an every day local agent. The great part of these models are the consistent speeds across all context ranges so when you pick a model that runs at 70tps, it generally stays within 10% of that up to 256k. The streaming packs are based on @turboderp_'s newly released self corrected packs that are optimized using the lossless DS4.1 as a teacher.
Posted by TheDavidTai (1.6k followers) 1 days ago · 33 likes · 1.7k views · view the original post on X. Kept by the AI Radar as Open & local models. Tools mentioned: deepseek-v4.1-flash-1017-streaming-repack-exl3.
More AI work like this
- Meet Huihui-Qwen3.6-35B-A3B-abliterated-MTP-GGUF: a 35B parameter image-text-to-text… — @HuggingModels
- An American AI 🇺🇸startup released a top decision 27B model. 👀 — @TeksEdge
- Qwen 4 is in my hands and I can't stop grinning 🔥 — @aijoey
- Qwen 4 (unnamed quant) — @loktar00
- Here's 4 steps to run an uncensored AI on your own GPU for $0: — @DeRonin_
- 北冥有鱼，其名为鲲。 — @NFT_Chen
- If you want to get started with Local AI but dunno where to start, THIS IS FOR YOU 🚨 — @javierbits
- 💥重磅！Qwen3.8-Flash-Next 量化版炸裂升级：Token消耗直接腰斩50%，准确率暴涨15%！本地32GB丝滑跑通125B级MoE！ — @NFT_Chen
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 65.3k posts from 5k X accounts over the last 14 days, 10.4k tools, 19 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-06 17:44 UTC. Full method.