Flint 40B decodes a code prompt 1.3x faster than Qwen3.8-27B on one RTX 5090. With 8…
This is a dev post classified by AGK Intelligence as AI dev tools (a tool drop), kept by the Dev Radar because it carries real work, not commentary.
Flint 40B decodes a code prompt 1.3x faster than Qwen3.8-27B on one RTX 5090. With 8 requests at once the order flips. RTX 5090, Flint 40B on llama.cpp: Decode, code prompt: 224.8 tok/s Decode, prose prompt: 224.9 tok/s First token, 32K: 5.7 s 1 at once: 193.9 tok/s 2 at once: 198.6 tok/s 8 at once: 198.5 tok/s RTX 5090, Qwen3.8-27B on NInfer: Decode, code prompt: 169.0 tok/s Decode, prose prompt: 150.1 tok/s First token, 32K: 3.6 s 1 at once: 148.3 tok/s 2 at once: 274.5 tok/s 8 at once: 711.9 tok/s Flint's recipe runs 1 slot, so extra requests wait in line, while NInfer's own line runs 2
Posted by Niklas (2.4k followers) 2 h ago · 0 likes · 234 views · view the original post on X. Kept by the Dev Radar as AI dev tools.
More dev work like this
- TensorFold 1.2.0 decodes Qwen3.8-27B code at 268.6 tok/s on an RTX 5090, level with 1.1.2. — @niklaslenz_ai
- DeepSeek V4.1 Flash 1017 for 4x DGX Sparks got a lot of updates, with support for… — @MiaAI_lab
- Vibe coding works when your agent has the right tools. — @apify
- 🔥必看！一个视频秒懂 DeepSeek Super-agent ！瞬间理解了梁圣的核心理念少占些注意力、多给些价值！ — @NFT_Chen
- Your coding agent says “done.” Where’s the evidence? — @DanKornas
- A $0.39 team of Qwen3.8 27B models just beat single call Fable 5 on a subset of… — @analogalok
- autonomous performance improvements, — @LukeParkerDev
- 🔥人气王！DeepSeek Harness 直接装进口袋，手机自己跑 Agent，再也不断档，电脑远程接着干！ — @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. 31.3k posts from 5k X accounts over the last 21 days, 8.9k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-01 10:02 UTC. Full method.