VQV Signal

ROBOTS & HARDWARE WATCH TECHNICAL

SGLang reports DeepSeek V4.1 Flash throughput rising to 1.09M tokens/sec per GPU on AgentX

According to a post on X by @sgl_project, the SGLang project says peak total throughput for DeepSeek V4.1 Flash on AgentX rose from 654K to 1.09M tokens per second per GPU over the past four days. The figure counts cache, new input and output tokens. For ultra-low latency, the post also reports 943 tokens per second per user at a concurrency of 1 using MegaKernel. The source is a social post and the snippet is truncated, so the full methodology is not visible here.

Source: Scobleizer News Pulse (X) · x.com Found via Robert Scoble’s Scobleizer News Pulse · a post on X, not independently verified Published 2026-10-31T04:35:00+00:00 Detected 2026-10-31T13:34:03+00:00
View original source

According to a post on X by @sgl_project, the SGLang project says peak total throughput for DeepSeek V4.1 Flash on AgentX rose from 654K to 1.09M tokens per second per GPU over the past four days. The figure counts cache, new input and output tokens. For ultra-low latency, the post also reports 943 tokens per second per user at a concurrency of 1 using MegaKernel. The source is a social post and the snippet is truncated, so the full methodology is not visible here.

AI-assisted summary based on the listed source.

SGLang says peak total throughput for DeepSeek V4.1 Flash on AgentX rose from 654K to 1.09M tokens per second per GPU over the past four days, counting cache, new input and output. For ultra-low latency, it also measured 943 tokens per second per user at a concurrency of 1 with MegaKernel. Original post on X...

Teams serving DeepSeek models on GPUs could get more output per card if these figures hold under their own workloads. The numbers are self-reported by the project in a social post, so reproducible benchmarks are the next thing to watch. It will also matter whether the AgentX and MegaKernel setups are documented well enough for others to test.

Hardware and robotics watchers may want to track whether this becomes a product, benchmark, or deployment signal.

Signal Strength 86% Technical label WATCH Public Interest 29 Category ROBOTS & HARDWARE Reader Depth TECHNICAL

Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.

Public Interest components
Recognizable Entity Score 0 Practical Impact Score 0 Novelty Interest Score 94 Consequence Score 30 Curiosity Score 16 Shareability Score 45

VQV surfaced this signal because it is recent, relevant to AI Chips, connected to Scobleizer News Pulse (X).