Open data · CC-BY 4.0 · DOI-backed

Nesyona Research

Nesyona publishes original research on the AI tool economy: real model API pricing, free-tier longevity, capability boundaries, and cost-per-outcome. Every dataset is first-party, permanently archived with a Zenodo DOI, and mirrored on Hugging Face and Kaggle. Part of the Lattice research network.

Published research

Dataset · 32 models tracked

AI Token Price Decay 2026: LLM API Cost Trajectory

The per-task API cost trajectory of frontier LLMs from March 2023 to 2026, normalized to one reference prompt so prices are directly comparable. Headline: ~1063x decay peak-to-floor, while Claude Opus held sticky pricing for 26 months against a collapsing market.

Study & findings Data (JSON) Zenodo DOI Hugging Face Kaggle

Cite: Couey, V. W. (2026). AI Token Price Decay 2026: LLM API Cost Trajectory [Data set]. Nesyona. https://doi.org/10.5281/zenodo.20632755

The Lattice research network

Nesyona is one node of the Lattice — a research-driven network of 17 open, DOI-backed, CC-BY datasets across 13 properties, all published under one ORCID-verified byline.
The network hub →All properties & the full research index Hugging Face →17 datasets, mirrored Kaggle →17 datasets, mirrored Lucreya ResearchAI GTM & GEO citation data Rinzara ResearchAI creative policy & rights data The authorVincent Couey · ORCID

All Nesyona datasets are released under CC-BY 4.0. Free to share and adapt with attribution to Nesyona / Vincent Couey (ORCID). Dated snapshots; verify current vendor pricing before relying on any value.

☆ Save
Dashboard