Thanks — you're asking which TTS and STT models work only with the TypeScript (TS) SDK and which work only with the Python SDK. • There are no models or plugins documented as TS-only. • Many STT and TTS providers are Python-only. STT vendors available only in the Python SDK include: Amazon Transcribe, AssemblyAI, Azure AI Speech, Azure OpenAI, Baseten, Cartesia (note: Cartesia TTS is cross‑SDK), Clova, ElevenLabs (STT listed for Python), fal, Gladia, Google Cloud, Mistral AI, Nvidia, Sarvam, Soniox, Speechmatics, Spitch. Deepgram and OpenAI STT are available in both Python and TS. • TTS vendors available only in the Python SDK include: Amazon Polly, Azure AI Speech, Azure OpenAI, Baseten, Gemini, Google Cloud, Groq, Hume, Inworld, LMNT, MiniMax, Nvidia, Sarvam, Smallest AI, Speechify, Spitch. The following TTS plugins are available in both Python and TS: Cartesia, Deepgram, ElevenLabs, Neuphonic, OpenAI, Resemble AI, and Rime. If you want, I can: • Produce a compact table of exactly which providers you plan to use, or • Help you pick a cross‑SDK option (e.g., OpenAI or Deepgram) and show example setup for TS or Python. Which would you prefer? Sources: Speech-to-text (STT) models | LiveKit Docs | Groq STT plugin guide | LiveKit Docs | Text-to-speech (TTS) models | LiveKit Docs | Models | LiveKit Docs | Agents (Node.js) | livekit/livekit_composite | DeepWiki