How Next AI Model works
Next AI Model lists every model release from the makers it follows, line by line, and shows where each line stands in its own release rhythm. It is checked for new releases every morning.
What counts as a release
A model is released on the first day anyone could use it: in an app or an API open to everyone, paid access and previews included, or as published weights. A model open only to partners, testers or a waiting list is not released yet. A model the maker has announced but not opened is shown as an announcement, with the maker's own words.
Lines
A line is one product, such as Gemini Pro or Claude Opus. A new product, a different tier (flagship, lightweight or coding) or a switch between open weights and closed starts a new line. Releases less than a week apart count as one launch. See every line we track.
How the status works
Each line's usual gap is the median of its last five gaps. The label says where the line stands in its own rhythm: Due from 85% to 100% of the usual gap, Past its usual gap after that, Mid-cycle in between, Just released for the first third. A line with fewer than three gaps between releases shows Too early to call instead: in our archive, "Due" on such lines was right no more often than chance.
Why no dates
We tested "usual gap ± 15 days" against every release in our archive: it would have caught fewer than 1 in 5 of them. AI release rhythms are too irregular for a date to be honest, so we show the rhythm, not a prediction. In our archive, a new model came within a month about 2 times in 5 when a line was reaching its usual gap, and less often once it was past it: that is why we never count "days late".
Quiet, Inactive and Discontinued
A line whose wait, measured in its own usual gaps, is longer than almost every past gap in our archive (fewer than 5 lasted that long) is marked Quiet. After a year without a release it becomes Inactive. Discontinued is only for lines the maker has officially closed.
Where the numbers come from
Model cards use the maker's own announcement and independent measurements. The Intelligence Index and the other scores marked "Artificial Analysis" come from Artificial Analysis, which tests AI models independently; blind-comparison scores come from Arena. Every card lists the pages its numbers were read from.
Disclaimer
Status labels are statistical readings of publicly available release dates, not insider information: actual release dates depend on development progress, strategic decisions and competition. Next AI Model is an independent project and is not affiliated with, endorsed by, or connected to any AI company mentioned on this site. This content does not constitute financial or investment advice. All trademarks and product names belong to their respective owners.
Questions or corrections: robfrata@gmail.com. What the site receives about you: Privacy.