Send a payload, see what the model says back. One model at a time. Speech in, text out — or the other way around.
Status: Pilot. Speech-to-text on all four tiers (Whisper, Bedrock Frankfurt, Mistral Voxtral). Text-to-speech on all four tiers (OpenAI, AWS Polly, Mistral Voxtral). EU Sweden customers: most Swedish requirements are satisfied by EU residency rather than strict Sweden-only residency — use EU Cloud.
Speech to text
Audio transcription needs a connected Brain Orchestra key. Connect yours to start transcribing — usage is billed to your own Brain Orchestra account.
Whisper, Bedrock Frankfurt, and Mistral Voxtral — billed per minute
EU Cloud and EU Strict transcription on EU-hosted infrastructure
Every model is available — there is no tier restriction