--- {"title": "How to follow new model releases as a machine learning student", "description": "Learn how machine learning students can follow new model releases, what to prepare, and how to compare tools that estimate release chances.", "url": "https://supakeeper.com/how-to-follow-new-model-releases-for-ml-students/", "date": "2026-11-03"} --- Learn how machine learning students can follow new model releases, what to prepare, and how to compare tools that estimate release chances. Follow new model releases by choosing the labs you care about, recording each release with its date, and checking your sources on a fixed routine. A tool such as Model Odds can add estimated release chances for Anthropic and OpenAI model families. [^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] | Read the figures as estimates, not dates. Each company page lists past releases. [^1] | | Lab announcement pages | Shows the posts each lab publishes about its own models. | Check the date and wording of each post yourself before recording it. | | General AI newsletter | Collects AI news in one place on a schedule you can read. | Check how often it publishes and whether it covers the labs you follow. | ## What you need before you start Start with a short list of the model families you want to follow, such as those from Anthropic and OpenAI, so you know which releases matter for your coursework. Keep a notes file or spreadsheet with the date you first saw each release and a link to its official post, so you can compare them later. You will also need a web browser. Model Odds is a web app that runs in the browser and needs JavaScript. [^1] If you want email alerts, have one email address ready that you can confirm. You can sign up with a confirmed email to get alerts. [^1] Decide how much time each week you will spend reading release news, since a routine works better than checking at random. Finally, pick a simple goal, such as following one release cycle or preparing a short summary for your study group. A clear goal makes it easier to decide which sources are worth keeping. ## Step by step 1. Choose the labs and model families you want to follow, and write them down. A short list is easier to keep up with. 2. Pick the sources you will check, such as the announcement pages of each lab and one news source you already trust. Two or three sources are usually enough for a student. 3. Create a simple record with the release name, the date you saw it, and a link to the official post. Dates kept in one place make later comparisons possible. 4. Set a fixed time each week to review your sources. A routine keeps the list current and is more reliable than checking at random. 5. If you want forward-looking estimates, add a tool that gives a chance of upcoming releases, then compare its estimates with the posts you have recorded. Treat each estimate as something to check, not as a fact. 6. Check whether a tool can replay past dates, so you can see how its estimates would have looked at an earlier time. This helps you judge how much weight to give them. 7. Share a short weekly summary with classmates or your study group. Explaining a release in your own words helps you notice gaps in what you have read. 8. Review the list each month. Drop sources you no longer read and add any new labs you need to follow. ## 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 of the two companies has its own page that lists every past release. [^1] Each past launch on the list has to be confirmed by an official post and an independent timeline, and it is listed with its date. [^1] Model Odds runs three estimators side by side and shows the spread between them as an error bar. [^1] The main estimate is a renewal process fitted on the gaps between recent launches, and it takes into account how long the current wait has already lasted. [^1] Signals read from company software kits and documentation, and from third-party model catalogues when two of them agree, can only raise the odds. [^1] For following releases over time, the email alerts and the public API may matter. Every subscriber gets every company's drops at the same time. [^1] The public API returns the odds, the launch record and service health. [^1] Model Odds is free, and so is its API, which needs no account. [^1] Other options include the announcement pages of each lab you follow, or a general AI newsletter you already read. Compare each option against the job of following new releases, and note which one gives you dates you can check. ## Common mistakes to avoid Treating an estimate as an announced date is a common error. An estimate is a chance, so check each official post before you record a release date. Relying on a single source can leave gaps. Keep at least two sources so one missed post does not leave you uninformed. Recording releases without dates or links makes later comparison hard. A short record with the date you saw each item helps with this. Counting unconfirmed reports as launches causes confusion. Model Odds lists only launches confirmed by an official post and an independent timeline, each with its date. [^1] Ignoring uncertainty is another mistake. Read the error bar as a sign of how far the estimators disagree with one another. [^1] Checking too often wastes time. A fixed weekly review is usually enough to keep up. ## Frequently asked questions ### Is Model Odds free to use? Yes. Model Odds is free, and so is its API, which needs no account. [^1] The website runs in the browser and needs JavaScript. [^1] Email alerts need a confirmed email address before they start. [^1] A student can therefore use the site and its API without paying. [^1] ### Does Model Odds give exact release dates? No. Model Odds gives estimated chances rather than dates. [^1] It ranks each model family by its estimated chance of a release within 30 days, so the figures are estimates and not announced dates. Compare any estimate with the official post once a company publishes one. The launch record lists past releases with their dates, which helps you check how the estimates have behaved. [^1] ### Can Model Odds send email alerts about new releases? Yes. You can sign up with a confirmed email to get alerts. [^1] Every subscriber gets every company's drops at the same time. [^1] Confirm the address you use, since alerts go only to a confirmed email. ### How does Model Odds make its estimates? Model Odds uses a renewal process fitted on the gaps between recent launches, and it takes into account how long the current wait has already lasted. [^1] It runs three estimators side by side and shows the spread between them as an error bar. [^1] Signals read from company software kits and documentation can only raise the odds. [^1] Reading the error bar alongside the main estimate gives a clearer picture of uncertainty. ### Can Model Odds test how its estimates would have done in the past? Yes, in a sense. The estimator can replay any past date without using data from after that date. [^1] This lets a student see how an estimate would have looked at an earlier point in time. Compare those replayed estimates with the verified launch record, which lists each past launch with its date. [^1] Together these checks can help you judge whether the method is worth trusting. Model Odds is free, and so is its API, with no account needed. [^1] - [Visit Model Odds](/go/start-modelodds) ## Sources [^1]: Model Odds' own description of itself (the product of this site's publisher) (/go/start-modelodds)