Skip to content
Trending storyDeveloping

SpaceX and DeepSeek GPU price figures compared in Tunguz analysis

1 article1 sourcesince Oct 20Last article

Overview

AISummary of 1 article

Tomasz Tunguz compares two GPU prices.

He says SpaceX's Q3 filing implies a negotiated price of $5.27 per card-hour, while DeepSeek claims a theoretical maximum of $38.00 per card-hour for its cheapest model. Tunguz says the inference business looks healthy at current prices. He says its case depends on demand persisting and on efficiency gains outpacing falling token prices.

Written by AI from the articles below · updated Oct 21, 1:16 PM ET

Check the sources:

Article timeline

The articles in this story. Times are ET.

Oct 20
  1. Tomasz TunguzBlog
    The GPU revenue ladder: inference pricing, SpaceX and DeepSeek figures

    AISpaceX's Q3 filing implies a negotiated GPU price of $5.27 per card-hour, while DeepSeek claims a theoretical maximum of $38.00 per card-hour for its cheapest model. Tunguz says the inference business looks healthy at current prices, but its case depends on demand persisting and efficiency gains outpacing falling token prices.

Heat Trend

Not enough continuous observations to show a trend yet.