Get AI model alerts by email
We watch every model release, API price change, and deprecation across the tracked labs so you don't have to check five different vendor pages. One email when something actually changes. Nothing on quiet weeks.
What we watch
New model releases, API price changes (input, output, cache-read), and deprecation or shutdown announcements across the tracked providers.
Where the data comes from
The same committed tracker data that renders the model directory, prices, and deprecations pages. No separate feed, no editorializing.
Cadence
Weekly, and only when something changed. If a week has zero tracked changes, you get zero emails that week.
A real week from the tracker
Week 36, 2026 (August 31, 2026 to September 6, 2026). This is exactly what a subscriber alert would have contained that week. Every line below is copied from the tracker's committed data.
Deprecations (7)
- gemini-robotics-er-1.6-preview (Google): shut down August 31, 2026, replacement gemini-robotics-er-2-preview
- gemini-omni-flash-preview (Google): shutdown announced for September 30, 2026, replacement gemini-omni-1.1-flash
- Assistants API (OpenAI): shut down August 26, 2026, replacement Responses API and Conversations API
- gpt-4o-mini-transcribe (OpenAI): shutdown announced for February 26, 2027, replacement gpt-live-transcribe or gpt-transcribe
- gpt-4o-transcribe (OpenAI): shutdown announced for February 26, 2027, replacement gpt-live-transcribe or gpt-transcribe
One email when a tracked model ships, changes price, or is deprecated. No noise, no ongoing cost.
Free tier
Everything above, for every tracked provider, by email. No price, no card required.
Planned: filtered alerts
A paid tier for alerts scoped to one provider or model, or programmatic access via a JSON API, is planned but not yet built or priced.