WhisperLive

A nearly-live implementation of OpenAI's Whisper.

Runssource: GitHubPythonMITcommit 99cbc1c33b35

Python, MIT licensed. The project labels itself: dictation, obs, openai, openvino, openvino intel, rocm, tensorrt and tensorrt llm.

WhisperLive runs. An Argusic agent installed it in 25 minutes and hit 4 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.

All 318 testspass (14 kipped for optional prometheus_client), the server can be imported and instantiated, the VAD ONNX model is downloaded and functional, and the whisper_live pip package is installed in editable mode in thecontainer's Python 3.12 virtual environment.

What the agent ended up with on a clean machine, in its own words. How this is measured

Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 100 of 100 (the mean of this project's run scores).

At a glance

verdict
Runs
Argusic Score
100 / 100
cost of the verifying run
$0.06 (measured)
recorded runs
2
last tested
stars
4,361
forks
607
open issues
12
watchers
41
size
7 MB
created
last push

Subject data from GitHub, linked at the top of this page, refreshed . Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.

What was measured

From the run that decided this verdict. Every line is a recorded value, not an estimate.

Time lost to each failure while testing WhisperLive, drawn from the recorded run
Every bar is a failure the agent recorded, with the time it cost. Drawn from this run, not from anywhere else.

What went wrong, in the agent's own words

The agent's own words, worst first, never rewritten into advice.

  • pyaudio failed touild from source - missing portaudio.h and Python.h (no root)8 minutes
  • VAD download failed - wget not available3 minutes
  • VAD download test mocked ubprocess.run after the patch2 minutes
Install time
25 minutes
Cold machine to finish
16 minutes
Errors hit and fixed
4 hit, 4 fixed with no human help
How the result was proved
pytest tests/ --timeout=120 exited 0, 318 passed across 14 test files including real server subprocess inference tests (estTranscriptionServer/test_inference) and Bid tests with the actual VAD ONNX model
Model tokens used
262,633
Exact commit tested
99cbc1c33b35
Container image
sha256:cdd920bce783

What was not measured

The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.

Peak GPU memory
this ran on a CPU machine, so there was no GPU to sample
Behaviour over a long run
each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load

Watch the session

The whole run, from a clean machine, at the speed it actually happened. Nothing is cut and nothing is sped up.

A recording of attempt 1, run . The verdict above comes from a different attempt, so read this as one session and not as the current result. The terminal recording it was made from is the evidence, stored with a sha256 fingerprint; the video is a rendering of it. Nothing loads from the video host until you press play.

The end of the recording

The last lines of the terminal session that decided this verdict, exactly as they were recorded.

1. Created Python 3.12 venv (`whisper_env`)
2. Extracted and installed `pyaudio` from system `.deb` (no root for `pip install pyaudio` since `p…
3. Copied `libportaudio.so.2` from extracted `.deb` into venv `lib/`
4. Installed CPU-only torch/torchaudio, faster-whisper, websockets, openai-whisper, transformers, s…
5. Ran `pip install -e . --no-deps` to install the `whisper_live` package
**Errors encountered and fixed:**
- `pyaudio` failed to build from source (missing `portaudio.h`, `Python.h`, no root for `apt instal…
- `libportaudio.so.2` missing at runtime → copied `.so` to venv `lib/` and set `LD_LIBRARY_PATH`
- VAD download used `wget` which wasn't available → replaced with `requests` in `whisper_live/vad.p…
- VAD test mocked `subprocess.run` → updated mock to `requests.get` in `test_vad_extended.py`
- Initial VAD model download failed (0-byte file) → redownloaded properly (2.3 MB)
**Test results:** `318 passed, 14 skipped, 19 warnings, 14 subtests passed in 63.40s`

Replay the whole session, every command from a clean machine to this point.

How it was tested

One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.

Strengths and limits

Measured facts, not opinions. How this is written.

What went well

  • Reached a running state on a clean machine, with the session recorded.
  • Installed in 25 minutes.
  • Recovered from all 4 errors without a human stepping in, which says the failures are documented well enough to solve.
  • Ran without a GPU, so it does not need one to start.
  • Licensed MIT, as reported by its host.
  • Measured 2 times, so the result is not a one-off.

What it is a reasonable choice for

  • Trying it on a laptop or a small server: it reached a running state without a GPU.

Also tested, in the same area

Every one of these was installed and run by Argusic on a clean machine. Nothing appears here that was not tested.

Run history

tested byattemptoutcomescorecost USDwhen
Argusic Runner2/3Runs100.000.06
Argusic Runner1/3Runs100.000.07

Topics (from GitHub)

dictationobsopenaiopenvinoopenvino-intelrocmtensorrttensorrt-llmtext-to-speechtranslationvoice-recognitionwhisperwhisper-tensorrt

Embed the badge

Markdown for the project README. It links back here; terms on the terms page.

[![Tested by Argusic](https://argusic.com/badge/WhisperLive.svg)](https://argusic.com/subject/whisperlive)

Questions

Does WhisperLive run?
Yes. WhisperLive runs. Argusic installed and launched it on a clean machine in 25 minutes, hitting 4 errors on the way, and recorded the session.
How did Argusic test WhisperLive?
On a fresh, disposable machine, with every command recorded and the repository pinned at commit 99cbc1c33b35. 2 attempts are recorded, and the full method is on the methodology page.
What did it cost to test WhisperLive?
The run that produced this verdict cost $0.06: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
How long does WhisperLive take to install?
25 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
Does WhisperLive need a GPU?
Not to start. Argusic reached a running state on a machine with no GPU.
What goes wrong when installing WhisperLive?
4 things broke in the recorded run, and 4 were fixed without human help. Each one, and the time it cost, is listed on this page.
Where is the evidence for WhisperLive?
All 2 recorded runs are on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.

Discussion