How to follow new model releases for AI hobbyists in 2026

Learn how AI hobbyists can follow new model releases from Anthropic and OpenAI, what to check in each option, and which mistakes to avoid.

By SupaKeeper · 2026-10-24

SupaKeeper is published by LunarWerx, which makes some of the tools compared here.

To follow new model releases, choose the companies you care about, pick one way to be told about changes, and check each item against an official post before you rely on it. Model Odds is one option that ranks Anthropic and OpenAI model families by their estimated chance of a release within 30 days. ^[1]

Option	What it does	What to check
Model Odds	Ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days. ^[1]	Check whether the two companies it ranks are the ones you follow. It is free, with no account needed. ^[1]
Company announcement pages	Shows each company's own posts about new models, read directly from the source.	Check how often you visit, since you will need to look for new posts yourself.
A general AI news source	Gives news about AI releases from a source you choose to read.	Check which labs it reports on and whether it gives the date of each release.

What you need before you start

Before you follow new model releases, decide which companies matter to you. Many hobbyists care about one or two labs, so write those names down to know what you are watching for.

Next, decide what kind of news you want. Some people want a quick heads-up when something appears, while others want a record they can look back through over time. Each need points to a different way of following releases.

Keep a place to note what you find, such as a notes file or a bookmark folder. A simple log of dates and model names makes it easier to spot patterns later.

You also need a way to be told about changes without checking every day. Decide whether you prefer email, a page you revisit, or a feed you read at your own pace.

Finally, read release news with some caution. Reports can be incomplete or change quickly, so note where each item came from.

Step by step

Write down the companies and model families you want to follow.

Find the official pages those companies use to announce models, and bookmark them.

Choose one way to be told about changes, such as an email alert or a page you check on a set schedule.

When something new appears, note the date, the model name, and the source you read it in.

Compare any claim against an official post before you rely on it.

Review your list each month and remove companies or sources you no longer follow.

Keep a short record of past releases so you can notice how often new models appear for each company.

Recheck your setup when your interests change, since the right way to follow releases depends on what you want to do with the news.

Tools that can do it

Model Odds is made by the team that publishes this site.

Model Odds ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days. ^[1] Each company has its own page that lists every past release. ^[1] Each past launch it uses has to be confirmed by an official post and an independent timeline, and it is listed with its date. ^[1]

It runs three estimators side by side and publishes the spread between them as the error bar, which is the range that shows how far the three estimates disagree. ^[1] The estimate is based on a model that looks at the gaps between recent launches and how long the current wait has already lasted. ^[1] Hints from company developer kits and documentation, and from third-party model lists when two of them agree, can only raise the odds. ^[1]

It can replay any past date without using data from after that date. ^[1] You can sign up with a confirmed email to get alerts, and every subscriber gets every company's releases at the same time. ^[1] Model Odds is free, and so is its API, with no account needed. ^[1] It is a web app that runs in the browser and needs JavaScript. ^[1] The website and the API share one snapshot that is cached for up to five minutes. ^[1]

Other ways to follow releases include reading each company's own announcement pages and using a general AI news source you choose to read. Compare each one against the job of following new model releases, and check how each shows its dates and sources.

Common mistakes to avoid

Avoid treating any estimate as a promised date. An estimated chance of release is a judgment about likelihood, not a schedule.

Avoid relying on a single post or an unconfirmed rumor. Check news against an official announcement before you act on it.

Avoid skipping the dates. A list without dates is hard to compare over time.

Avoid following too many sources at once. A short list you actually read is more useful than a long one you ignore.

Avoid forgetting to recheck your setup. Interests change, and a routine that worked last month may need small changes.

Frequently asked questions

Does Model Odds cost anything to use?

Model Odds is free, and so is its API, which lets other programs read the data, with no account needed. ^[1] It is a web app that runs in the browser and needs JavaScript turned on. ^[1] It is also a website with a public API, so you can read its odds without signing up. ^[1] Check that your browser runs JavaScript before you start, since the page will not work without it.

Can Model Odds send alerts about new releases?

You can sign up with a confirmed email address to get alerts. ^[1] Every subscriber gets every company's releases at the same time. ^[1] Signing up is free, and no account is needed to read the site itself. ^[1] After you sign up, check your alert settings so the messages arrive where you want them.

How does Model Odds work out its estimates?

It runs three estimators side by side and publishes the spread between them as the error bar. ^[1] The estimate is based on a model that looks at the gaps between recent launches and how long the current wait has already lasted. ^[1] Hints from company developer kits and documentation can only raise the odds, and the same goes for third-party model lists when two of them agree. ^[1] Keep in mind that the error bar shows disagreement between the estimators, not a promised date. ^[1]

Can Model Odds check its past estimates against real history?

It can replay any past date without using data from after that date. ^[1] Each past launch it uses has to be confirmed by an official post and an independent timeline, and it is listed with its date. ^[1] You can read the per-company pages to see every past release for Anthropic and OpenAI. ^[1] That gives you a starting point for checking the record yourself.

Does Model Odds offer data that programs can read?

Its API serves the odds for each model family, the launch record, and service health, for programs that want to read them. ^[1] The website and the API share one snapshot that is cached for up to five minutes, so the numbers can be up to five minutes older than the latest change. ^[1] Check the timestamp on any figure you copy into your own notes.

Model Odds is free, and so is its API, with no account needed. ^[1]

Try Model Odds

Sources

Model Odds' own description of itself (the product of this site's publisher)