Nvidia bought the open model commons. OpenAI published its retirement calendar.

Take the commitments at face value and the structural point still stands: the default distribution channel for open models is now owned by the company that sells the hardware they run on. Nothing breaks this week. What changes is that "we can always self-host an open model" is now a dependency on one vendor's continued goodwill rather than on a neutral commons. Worth knowing which of your fallbacks actually resolve to huggingface.co.

Every one of these was announced months ahead, in public, with a date. The failure mode is never the notice, it is that nobody owns checking. Put the page in a calendar reminder the way you would a certificate expiry, because that is functionally what it is.

Every other item on that page offers a migration target. This one does not, which makes it a vendor selection problem rather than a version bump. If video generation is in your product, you have seventeen days to pick a different provider and re-test output quality, not seventeen days to change a model string.

Five months is a generous window. The risk is not the deadline, it is that Whisper is typically wired into a background job written once, and background jobs fail into a queue rather than into a user's face. Grep now, migrate on your own schedule, rather than in February.

Two shutdowns in a fortnight where the product ends and the team lands somewhere larger. If a workflow automation tool sits between two of your systems, the question is not whether it is well funded, it is how long your integration keeps running after an announcement and whether you have an export.

That is roughly 86 times revenue, which tells you the price is for the distribution position and the developer graph, not the P&L. A useful benchmark if you are raising on a developer-adoption story rather than a revenue one.

Voice input keeps attracting late-stage money while the underlying transcription APIs churn. If you are building on voice, note that the capital is flowing to the interface layer, not to the model layer you would be renting.

These are the pieces teams build process around rather than product: your evaluation history and your prompt library live here. Export before the date, because a shut-down dashboard takes your baseline measurements with it and you cannot re-run last quarter's evals afterwards.

If your architecture assumes an open-weight escape hatch, price in that the escape hatch may be owned by a strategic buyer within a year. Portability is a property of your code, not of the model licence.