jiwer
Evaluate your speech-to-text system with similarity measures such as word error rate (WER)
Runssource: GitHubPythonApache-2.0commit 2227002889f2
Python, Apache-2.0 licensed. The project labels itself: automatic speech recognition, evaluation metrics, python3, speech to text, wer and word error rate.
jiwer runs. An Argusic agent installed it in 0.5 minutes and hit 3 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.
All 82 tests pass and the jiwer CLI and Python API both work correctly, computing word error rate and related ASR evaluation metrics using RapidFuzz C++ backend.
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
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.

What went wrong, in the agent's own words
The agent's own words, worst first, never rewritten into advice.
- Externally-managed Python environment (Debian/Ubuntu) prevented system-wide pip install0.2 minutes
- dev dependency group uses uv-specific pyproject.toml format [dependency-groups] not recognized by pip, so pytest, ruff, pytest-benchmark not installed0.1 minutes
- test_speed.py uses pytest-benchmark fixture but it wasn't installed0.1 minutes
- Install time
- under a minute
- Cold machine to finish
- 1 minutes
- Errors hit and fixed
- 3 hit, 3 fixed with no human help
- How the result was proved
- pytest tests/ -v exited 0, 82 passed including alignment, CER, WER, empty-ref, large-vocab, speed, and transform tests; jiwer --help printed CLI usage; Python API jiwer.wer('hello world', 'hello duck') returned 0.5
- Model tokens used
- 46,039
- Exact commit tested
- 2227002889f2
- 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
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
test_speed_n1 15.0912 (1.0) 267.3490 (1.0) 16.3751 (1.0) 4.9650… test_speed_n10 87.6330 (5.81) 400.2841 (1.50) 92.1648 (5.63) 7.9659… test_speed_n100 844.9689 (55.99) 8,068.0160 (30.18) 889.0082 (54.29) 440.2493… test_speed_n1000 8,419.3682 (557.90) 17,442.2171 (65.24) 9,237.0528 (564.09) 2,131.8298… Legend: Outliers: 1 Standard Deviation from Mean; 1.5 IQR (InterQuartile Range) from 1st Quartile and 3rd… OPS: Operations Per Second, computed as 1 / Mean ============================= 82 passed in 11.72s ============================== All 82 tests pass. Everything works perfectly. tokens used 46,039 All 82 tests pass. Everything works perfectly.
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 0.5 minutes.
- Recovered from all 3 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 Apache-2.0, 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.
- A quick evaluation: it was running 0.5 minutes after the clone.
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
Topics (from GitHub)
automatic-speech-recognitionevaluation-metricspython3speech-to-textwerword-error-rate
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/jiwer)Questions
- Does jiwer run?
- Yes. jiwer runs. Argusic installed and launched it on a clean machine in 1 minutes, hitting 3 errors on the way, and recorded the session.
- How did Argusic test jiwer?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit 2227002889f2. 2 attempts are recorded, and the full method is on the methodology page.
- What did it cost to test jiwer?
- The run that produced this verdict cost $0.02: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does jiwer take to install?
- 0.5 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 jiwer need a GPU?
- Not to start. Argusic reached a running state on a machine with no GPU.
- What goes wrong when installing jiwer?
- 3 things broke in the recorded run, and 3 were fixed without human help. Each one, and the time it cost, is listed on this page.
- Where is the evidence for jiwer?
- 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.