Observed arrival · 2026-10-13
Inferenstack’s labeled map of AI inference costs
A free, no-sign-in set of boards compares model API prices, GPU rental rates, and the costs of buying, renting, or building inference capacity.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- AI infrastructure buyers and model builders
- Worth noticing
- Its labels distinguish measured observations, formula-based estimates, and unverified vendor-published figures; speed and reliability are not probed.
Field notes
The boards separate provider-reported prices from estimates and measured observations, and include comparisons such as model costs across providers and on-demand GPU rates. A GPU-fit example names an RTX 4090 with 24 GB, while a deployment example models 25 people and shows an estimated ±50% accuracy band. The homepage reports an archive of 68,915 price observations, but says speed, reliability, and rate limits are not covered.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
ƒJavaScriptBrowser-side code central
One card from the complete issue