DeepSeek serves 150T tokens/day on one cluster, beating four inference clouds combined
natolambert · x · 2026-10-24
Nathan Lambert relays DeepSeek's disclosure: it serves 150T tokens per day on a single cluster — more open-model inference than OpenRouter, Fireworks, Together, and BaseTen combined. He also reconstructs comparative inference-volume estimates across those providers, a hard benchmark for the open-model serving ecosystem.
More from Infra
- PS5 Pro sells out in Toronto as memory shortage hits consumer hardware — Rasmic · 2026-10-24
- Nvidia compute unlock boosts RTX 3080 20GB prefill by up to 29%, decode unchanged — No-Refrigerator-1672 · 2026-10-24
- Conflicting RSI evidence: NVIDIA says AI-assisted CUDA optimization went 5x, then 2x in a month — soumitrashukla9 · 2026-10-24
- Miles v0.1.3 cuts first RL weight update from 99.5s to 3.9s, shares GPUs across train/rollout — hsu_byron · 2026-10-24
- 98%+ Tensor Core utilization shown in LLM training run, drawing admiration — gharik · 2026-10-24
- We spent millions on AI apps with zero revenue — but our old GPUs are now worth more than we paid — stephen280up · 2026-10-24