# collabora/WhisperFusion

WhisperFusion builds upon the capabilities of WhisperLive and WhisperSpeech to provide a seamless conversations with an AI.

Repository: https://github.com/collabora/WhisperFusion
Canonical: https://ross.abutalabs.com/products/whisperfusion
Language: Python
License Family: other
Last push: 2024-07-31T13:28:41+00:00

## Health v2 (maintenance only)
Score: 26/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 70
- inputs: {"age_days": 992, "days_push": 763, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1647, forks 126 (observed 2026-08-28T04:05:16.520539+00:00)

## What it is
WhisperFusion is a real-time voice chat application that combines WhisperLive speech-to-text, a Mistral/Phi LLM, and WhisperSpeech text-to-speech for ultra-low latency AI conversations. It runs the models as optimized TensorRT engines on a single NVIDIA GPU with a web-based GUI.

## Use cases
- talk to an AI assistant by voice with ultra-low latency
- build a real-time speech-to-speech chatbot
- run a local voice assistant on an RTX 4090 GPU
- demo low-latency LLM voice conversations
- integrate Whisper transcription with an LLM and speech synthesis

## When to choose
- you need real-time voice conversations with an AI on NVIDIA GPUs
- you want a self-hosted speech-to-text + LLM + TTS pipeline
- you need minimal latency between speaking and hearing a response

## When to avoid
- you lack a 24GB+ NVIDIA GPU
- you need a production-ready, licensed product
- you want cloud-hosted or CPU-only voice assistants

## Facets
- artifact type: application
- maturity: experimental
- function: speech-recognition, llm-inference, tts, chatbot, gpu-computing
- domain: artificial-intelligence, large-language-models, speech-processing
- platform: python
- tags: voice-assistant, tensorrt, real-time-conversation, whisper, low-latency, real-time, linux, gpu, docker, web-server

## Member repositories
- collabora/WhisperFusion (main) score 26

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:16.520539+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:45:17.987074+00:00, confidence not recorded.
  - readme: https://github.com/collabora/WhisperFusion (fetched 2026-08-28T04:05:16.520539+00:00, sha 9eda9bd0b037)
- Data as of 2026-08-30T08:39:29.467469+00:00.