GLM-5.3-Flash, a 321B MoE, decodes 16.0 tok/s on one DGX Spark with a 128K window
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.
GLM-5.3-Flash, a 321B MoE, decodes 16.0 tok/s on one DGX Spark with a 128K window 1 DGX Spark, 128 GB, Project Maya v1.1.53 Decode, code: 16.0 tok/s Decode, prose: 16.0 tok/s 500 words (code and prose): about 43.2 s Prefill at 8K: 689 tok/s First token at 32K: 46.9 s First token at 128K: 193.4 s Decode with a 128K prompt: 14.5 tok/s Quant: Maya-S IQ2_XXS, 96.5 GB, about 2.4 bpw All 12,096 experts fit in the Spark's memory pool Setup: the author's installer, 131,072 window as in the README's Spark row, 1 slot, MTP drafting Speed test: decode median of 2 runs, prefill of 3 or 4, 512 tokens, te
Posted by Niklas (2.8k followers) 18 h ago · 27 likes · 3.6k views · view the original post on X. Kept by the AI Radar as Open & local models. Tools mentioned: niklaslenz.com.
More AI work like this
- Qwen 4 (unnamed quant) — @loktar00
- 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
- This could be so cool. — @niklaslenz_ai
- Qwen is teasing a new model — @RoundtableSpace
- 时代变了，以前是token每秒，有人提出了秒每token。。。。 — @mylifcc
- Qwen3.8-Flash-Next GGUFs now run faster with less thinking and less token use. — @UnslothAI
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. 65.6k 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 16:05 UTC. Full method.