deepseek-v4.1IMAGE
DeepSeek V4.1 Flash hits 75,000 tok/s output on a B300x8 machine
A user reports running V4.1 Flash as the backbone for agent fleets on a B300x8 machine, reaching 75,000 tok/s output.
Justin Waugh
@JustinWaughArchived 12 days agoOriginally posted 2026.10.18
Justin Waugh @JustinWaughOriginally posted 2026.10.18
75,000 tok/s output feels wild. Deepseek v4.1 flash as a backbone for fleets on a B300x8 machine is amazing.
x.com/justinwaugh/status/2111664122542563693Archived 2026.10.18 16:31