Audio transcription with OpenAI Whisper and GPT-4o, plus audio format conversion and text to speech.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add whisper --env OPENAI_API_KEY='<OPENAI_API_KEY>' --env AUDIO_FILES_PATH='<AUDIO_FILES_PATH>' -- uv sync~/.codex/config.toml
[mcp_servers.whisper]
command = "uv"
args = ["sync"]
env = { OPENAI_API_KEY = "<OPENAI_API_KEY>", AUDIO_FILES_PATH = "<AUDIO_FILES_PATH>" }
# codex mcp add whisper --env OPENAI_API_KEY='<OPENAI_API_KEY>' --env AUDIO_FILES_PATH='<AUDIO_FILES_PATH>' -- uv sync~/.cursor/mcp.json
{
"mcpServers": {
"whisper": {
"command": "uv",
"args": [
"sync"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"AUDIO_FILES_PATH": "<AUDIO_FILES_PATH>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"whisper": {
"type": "stdio",
"command": "uv",
"args": [
"sync"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"AUDIO_FILES_PATH": "<AUDIO_FILES_PATH>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"whisper": {
"command": "uv",
"args": [
"sync"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"AUDIO_FILES_PATH": "<AUDIO_FILES_PATH>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"whisper": {
"command": "uv",
"args": [
"sync"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"AUDIO_FILES_PATH": "<AUDIO_FILES_PATH>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"whisper": {
"command": "uv",
"args": [
"sync"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"AUDIO_FILES_PATH": "<AUDIO_FILES_PATH>"
}
}
}
}Environment variables
OPENAI_API_KEYAUDIO_FILES_PATH
Replace the placeholders with your own credentials and keep them out of version control.