Where do the models run?

Chat, speech and transcription models run on Together AI's serverless GPU fleet in the United States, behind our gateway (authentication, metering, aliasing, error normalisation). OmniParser runs on our own hardware. If you need a specific region, contact us.