About Next AI Drop

What is Next AI Drop?

Next AI Drop is a forward-looking AI model release timeline that fuses Polymarket and Kalshi prediction-market odds, hourly intel, and a Drop Readiness score onto one forecast card. Instead of logging launches after they happen, it estimates which frontier models, IDEs, and agents are about to ship — and how confident that estimate is — so builders can plan ahead of a release rather than react to it.

Each forecast card carries four things on a single line: a blended odds reading from two markets, the freshest tagged intel, the decoded internal codename where one exists, and a Drop Readiness number that compresses all of it into one figure. The timeline itself runs from the recent past through today and out into named forecast windows, which is the part no other tracker draws.

The thesis: a timeline that points forward

Every other AI release tracker is a record of what already shipped; Next AI Drop is the only one whose timeline extends past today into forecast windows. Calendars and launch logs are rear-view mirrors — accurate about the past, silent about next week. We treat the next drop as the thing builders actually care about and put a date range and a confidence score on it.

That forward view is the whole point. Visit the live forecast board and the entries past today are not guesses pulled from the air — they are anchored to market consensus and tagged intel, ranked by Drop Readiness so the most imminent drops sort to the top. The category truth is that trackers tell you what landed. The twist is that we tell you what is about to.

Who it is for

Next AI Drop is built for AI builders, engineers, and teams who plan their work around frontier drops. If your roadmap depends on whether a new flagship model ships this month or next quarter — which provider gets there first, which context window or price point to design against — you are who this is for. It is a planning instrument, not a news feed.

The audience is deliberately narrow. This is not a consumer hype site or a benchmark leaderboard. It is for the person deciding whether to wait two weeks for a confirmed release or build against today's model, and who wants that decision grounded in a forecast rather than a rumour thread.

What makes it different

Next AI Drop combines four things no single competitor puts together: cross-venue odds from Polymarket and Kalshi, hourly tagged intel, codename decoding, and a Drop Readiness score on one card. Prediction markets give the odds, the intel layer gives context and recency, the codename map links pre-release SKUs to the public model they likely become, and Drop Readiness collapses the lot into one comparable number.

The score is computed the same way everywhere it appears: DR = .45 odds + .25 intel + .20 deadline + .10 volume. Odds carry the most weight because two markets pricing the same event is the strongest public signal we have; intel recency, deadline proximity, and market volume fill in the rest. The full computation, the stages (imminent, expected, rumoured) and the intel tags (Confirmed, Rumor, Market) live on the page that owns them — read how Drop Readiness is calculated. To see how internal names like ⟨MYTHOS⟩ or ⟨EMBER-ALPHA⟩ map to public releases, see the codename watch.

How we sit between markets and trackers

Next AI Drop is the synthesis layer between prediction markets and editorial release trackers. Prediction markets price the odds of a release but carry no editorial context or multimodal breadth; editorial release trackers list dates and rumours but refresh on a multi-day cadence and hold no market data. We take both, add the missing venue and the missing layer, and put the result on one card.

Polymarket and Kalshi are sources we build on, not rivals — we read their odds and add the other venue, hourly intel, and a Drop Readiness score on top. Against editorial trackers the difference is freshness and scope: our intel is cached hourly, while published competitor cadences run slower. For an honest, feature-level breakdown of how we compare to release trackers, see the compare hub.

Our values

Next AI Drop runs on three values: signal not stakes, source-linked, and free. Together they define what the product will and will not do — a forecasting tool that never becomes a casino, a claim register where every figure is traceable, and a public utility with no paywall, no ads, and no paid placement.

• Signal, not stakes. Prediction-market odds are used as a forecasting signal only. Next AI Drop is not betting, gambling, or financial or investment advice, and nothing here is a recommendation to trade — see the markets are signal, not stakes note.
• Source-linked. Forecasts trace back to the market and intel behind them; we explain method and ranges rather than asserting numbers without provenance. More on our data sources.
• Free. No paywall, no ads, no paid placement, no sponsored rankings. The forecast board is the same for everyone, and no provider can buy a higher Drop Readiness score.

Who is behind it

Next AI Drop is an independent project operated as a solo effort by Wibo van der Sluis, based in Amsterdam, Netherlands. It is not backed by, affiliated with, or influenced by any AI lab, prediction market, or tracker — including Polymarket and Kalshi. Independence is the reason the rankings can stay free of paid placement and the comparisons can stay fair.

One person operating the project keeps the editorial line simple: the only incentive is to forecast the next drop accurately, because that is the entire value of the site. There is no partnership to protect and no advertiser to please.

Contact

Reach the operator at hello@nextaidrop.com for corrections, questions, and source tips. Spotted a wrong date, a mislabelled codename, or a forecast that drifted from reality? Send it — corrections are the fastest way to make the timeline better, and source tips help the forward view stay sharp.

For ongoing updates, subscribe to the RSS feed. It carries the same hourly-refreshed intel that powers the forecast board, so you can track what is about to ship without opening the site.