a16z's Charts of the Week for Oct 30 (by Moses Sternstein) looks at the "Rack Spread": GPU rental prices are rising while token prices fall. The official a16z X account summed it up on Oct 31: since May, average token prices are down 67% and token spend is up more than 120%. By the article's napkin math, Ramp data gives roughly 122%, and Ornn's blended price multiplied by OpenRouter volume gives roughly 160%. A separate post says that since mid-July B300 rentals are up 106%, B200 up 66% and H100 up 17%. a16z also stresses that these indexes come from small, biased samples and do not show true total demand.
Key Takeaways
- ✓Prices and spend: since May, average token prices are down 67% and token spend is up more than 120% (a16z post on X, a16z figures)
- ✓Two napkin estimates: Ramp token-spend data shows about +122%; Ornn blended price times OpenRouter volume shows about +160% (Charts of the Week, described by the author as napkin math, not a rigorous method)
- ✓GPU rentals since mid-July: B300 +106%, B200 +66%, H100 +17%, while average token prices fell 58% (a16z post)
- ✓AI spend per employee: an Atlanta Fed survey puts the mean above $1,000, while Ramp's median is $15.70 a month ($188 a year)
- ✓Reach: the two posts have about 371 and 492 likes and 38K and 96K views (as of 18:15 SGT)
Key Decision Metrics at a Glance
Turn your technical choice into a development budget
Compare 40 dev plans & simulate token costs vs $20/mo subscriptions
Project Links & Resources
Direct AccessIn-Depth Technical Analysis
Background
Around July, a chart went viral showing GPU rental prices rising while token prices fell. Some called it the "Crunch Spread", after the oil industry's crack spread; a16z editor Alex Danco renamed it the "Rack Spread". The bearish reading is that tokens are becoming a commodity, squeezing model-layer margins and making compute spending hard to pay back. a16z's Oct 30 Charts of the Week responds to that reading.
How the argument works
The core point is that oil and refined products are bound by volume, but there is no fixed conversion between GPUs and tokens. When token prices fall, volume can grow faster, so total spend rises. The official a16z X account summed it up on Oct 31: since May, average token prices are down 67% and spend is up more than 120%.
Data comparison
This is not a model release, so there are no benchmarks. The article gives two napkin estimates: Ramp data shows about +122%, and Ornn's blended price times OpenRouter volume shows about +160%. On GPUs, since mid-July B300 rentals are up 106%, B200 up 66% and H100 up 17%. The author warns that Ramp, OpenRouter, Silicon Data and Ornn all draw on small samples of users who expose their usage, which mostly excludes large enterprises and skews toward cost-focused users who route to cheaper models. All figures are a16z's.
What it means for developers
A lower unit price does not mean a lower bill, and multi-step agent tasks multiply token usage. Track token use per task, not just the unit price; cap long-running tasks; and look at model routing and caching. Treat these numbers as directional, not as the true industry total.
Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.
Discussion & Comments
0Sign in to join the discussion
Connect with AI developers to exchange benchmark insights.