2 months ago
When will railway be able to offer GPU instance support? Particular to run vLLM instances on demand for inferencing. I have seen a few threads that are old on this subject and obviously cant run those type of solutions on CPU machines. This seems like very important feature or will have consider other hosting solutions.
1 Replies
2 months ago
This thread has been opened as a bounty so the community can help solve it.
Status changed to Open Railway • about 2 months ago
2 months ago
On May 20, 2026, Railway's CEO Jake Cooper gave this answer on an interview when asked about GPUs:
"We will not do GPUs now, but we 100% will do GPUs at some point in the future."
And since he was experimenting with it back in 2024 and stopped and then said this, I can tell u that most likely it's not going to be this year.
For an actual ETA of when it is going to be offered, no one can tell, not even the moderators as on Discord he has denied planning on GPUs again recently, so he is not currently working on it.