DeepSeek V4.1 Flash inference speeds up 1.5x on GB300, up to 5x on Vera Rubin

Artificial Analysis' AA-AgentPerf benchmark shows DeepSeek V4.1 Flash got 1.5x faster on GB300 within a month, while NVIDIA's updated Vera Rubin NVL72 delivers up to 5x per-GPU throughput over GB300.

2026-11-07 ~ 2026-11-07 · 2 related posts

• DeepSeek V4.1 Flash gets up to 1.5x faster on GB300 NVL72 in one month — ArtificialAnlys · 2026-11-07
• Vera Rubin NVL72 hits 5x GB300 per-GPU throughput on DeepSeek V4.1 Flash — ArtificialAnlys · 2026-11-07