Reachy Mini's conversation software can use a voice-processing system on a computer you control. The local conversation tutorial connects four stages: detecting speech, transcribing it, generating a reply and converting the reply into audio. A laptop can run those stages while the robot connects over the local network, so “local” does not require every model to run inside the robot. Keeping every stage local avoids sending conversation audio to a hosted inferenceRunning a trained model to get answers — the phase where AI is actually used, as opposed to trained.Full definition → service. The same software also supports a remote LLMA large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.Full definition →, so that privacy property depends on the selected configuration. Separate stages let owners replace individual models as their needs change.
Terms in this piece · Glossary
inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
