VoiceLab · MCP
Connect the hosted VoiceLab Model Context Protocol (MCP) server to ChatGPT, Cursor, or another MCP client. It lets an AI client run speech synthesis and transcription through VoiceLab.
Add a remote MCP server in your client and enter this address as its endpoint:
MCP endpoint
https://mcp.voicelab.uz/mcpSign in with your VoiceLab account through OAuth. The MCP server uses VoiceLab OAuth. The X-Api-Key documented for the Aisha API is not an authentication method for this endpoint.
Client setup instructionsUse MCP when a task should happen inside an AI client conversation and the client should choose from available VoiceLab tools. text_to_speech creates audio from text; speech_to_text starts an audio transcription. list_voices and list_tts_languages help discover available options.
Use the Aisha API references when you need server-to-server integration from your application, explicit request and response control, or a real-time audio stream. These API routes use Aisha API authentication and their own endpoints; that is separate from signing in to the MCP server.
The MCP client discovers tools from the server. For TTS, check available languages and voices in the client. The STT tool starts a job from an audio source and returns it asynchronously; check its status with get_transcription until the result is ready.