Microsoft and OpenAI dissolved their exclusive partnership and revenue-sharing arrangement, freeing OpenAI to use any cloud or compute vendor.
Key Takeaways
The exclusive deal locked OpenAI to Azure infrastructure; its removal lets OpenAI now evaluate AWS, GCP, and custom silicon.
Revenue-sharing between Microsoft and OpenAI is terminated under the restructured arrangement.
Microsoft publicly framed the change as a “simplification grounded in flexibility,” not a breakup.
OpenAI’s API workloads and model training are no longer Azure-bound, opening multi-cloud options immediately.
Hacker News Comment Review
Commenters broadly see Google as the biggest structural winner: with Azure exclusivity removed, OpenAI can now adopt Google’s Gen 8 TPU, previously off-limits.
Strong consensus that the terms heavily favor OpenAI; multiple commenters questioned why Nadella accepted this as Microsoft’s equity stake fell from roughly 49% to 27%.
One commenter documented the press release being rewritten mid-cycle: the original version retained Azure API exclusivity language and AGI framing; the revised, shorter version removed both.
Notable Comments
@ZeroCool2u: argues Google Cloud may now be the only hyperscaler able to resell all three major frontier lab models, raising the value of GCP credits.
@concinds: “Am I crazy, or was this press release fully rewritten” – documents specific removals including AGI language and contradictory Azure API exclusivity claim.
