CueLight vs Pipecat

CueLight

CueLight gives a voice pipeline a local end-of-turn probability every 20 ms from audio alone, with a 1.9 MB model and Apache-2.0 weights. It is one day old, English only, installs from Git rather than PyPI, has no CI or changelog, and its accuracy figures are the author's own.

Pipecat

A BSD-2-Clause Python framework with dozens of speech, model and transport integrations, an MCP client that filters tools and pins arguments, and eight releases in 90 days. It has no tool-approval step, and 1.12.0 can send OPENAI_API_KEY to other OpenAI-compatible providers when no key is passed, a flaw fixed on main but not yet released.

CueLight

1. Feed 320 float32 samples at 16 kHz mono per push. Resample first, or the call raises an AssertionError
2. Spell the preset as snappy or careful. Any other name silently uses default thresholds
3. Call clear() before a new call or session. Pass reset_on_turn_start=False if your pipeline already resets per turn
4. Pin weights by downloading them to a local folder and passing that path to from_pretrained
5. Don't use it to identify speakers or clone voices. The training data's conditions forbid it

Pipecat

1. Pass api_key explicitly to every OpenAI-compatible service such as Groq, Cerebras or OpenRouter. On 1.12.0 a missing key sends OPENAI_API_KEY to that provider
2. Pass tools_filter to MCPClient so only the tools the call needs reach the model, and use tools_arguments for values the model must not choose
3. Gate tools that book, pay or send in your own handler. Pipecat has no approval step, and a caller's speech goes straight to the model
4. Keep the development runner on localhost. Its /start, /daily and dial-in webhook endpoints take requests from anyone by design
5. Pin a minor version and read the breaking-change flags in CHANGELOG.md before upgrading, since minor releases remove and rename APIs