Nemotron 3.5 Bolt 74B decodes prose at 265.5 tok/s on one RTX 5090 and 72.5 on one DGX…
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.
Nemotron 3.5 Bolt 74B decodes prose at 265.5 tok/s on one RTX 5090 and 72.5 on one DGX Spark RTX 5090, 32 GB Decode, prose: 265.5 tok/s Decode, code: 219.8 tok/s 500 words (code and prose): about 2.8 s First token at 128K: 18.7 s 1 DGX Spark, 128 GB Decode, prose: 72.5 tok/s Decode, code: 63.4 tok/s 500 words (code and prose): about 10.2 s First token at 128K: 67.8 s Experts at 2.06 bpw (IQ2_XXS), LiLiCorr drafts, 262,144 context Speed test, median of 2 runs, 512 tokens, temperature 1.0, top_p 0.95, thinking off To fit 32 GB and avoid a known crash at 8 slots, the 5090 runs 2 slots instead
Posted by Niklas (2.7k followers) 2 h ago · 3 likes · 419 views · view the original post on X. Kept by the Dev Radar as AI dev tools.
More dev work like this
- Stop rebuilding memory and tool wiring for every agent. — @DanKornas
- 😱卧槽！万物皆可逆向！Rust神器craft-anything炸场，看懂任何闭源软件，AI直接帮你重造一个自己的版本！ — @NFT_Chen
- I look at the GBrain Tools screen quite a bit to see what all my agents are up to during… — @bradgessler
- ComfyUI 玩家这回能直接抄作业了，有人把 MiniMax H3 的“导演台”做好了。 — @bkdgiffug
- Just hit 238 followers — thank you 🙏 — @mims3590
- Introducing SubQ Snap 1.1 on Baseten. — @baseten
- Meet Qwen3.5-4B-GLB-3D-Coder: a 4B model that turns text into 3D GLB files and Three.js… — @HuggingModels
- I run OpenCode inside T3 Code. — @theayush
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, 9.1k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-02 21:12 UTC. Full method.