texterrors
For WER
Decision gist · record as of 2026-08-14
Yes. texterrors is actively maintained, has no known vulnerabilities, carries a permissive license, and directly solves a common problem in speech recognition evaluation. Medium install friction is acceptable given the prebuilt wheels and straightforward dependency set. Install it if you regularly evaluate ASR or transcription systems.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.9 or newer; compiled dependencies (Levenshtein, numpy) may require build tools on some platforms.
- Medium install friction due to compiled dependencies (Levenshtein, numpy, regex).
- Active maintenance with recent release (66 days ago).
License · maintenance · safety
Apache-2.0 License (permissive) — Apache-2.0 permissive license allows use in commercial and proprietary projects with minimal restrictions.
last release 2026-06-09 (66 days) · last repo commit 2026-06-09 · 37 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 875,922 downloads/mo, #4,832 on PyPI
Alternatives
Verify before relying
pip install texterrors
from texterrors import align_texts
ref_aligned, hyp_aligned, cost = align_texts(
["speedbird", "eight", "six", "two"],
["hello", "speedbird", "six", "two"],
use_chardiff=True,
)- Whether the character-aware alignment mode is enabled by default or requires explicit flags
- Performance characteristics on very large transcription files or datasets
- Compatibility with non-English or multilingual transcription outputs
What it is and what it does
texterrors is a command-line tool and Python library for evaluating automatic speech recognition and transcription systems. It computes standard metrics like word error rate (WER) and character error rate (CER), then provides detailed aligned error reports showing exactly where and how transcriptions diverged from reference text. The tool supports multiple input formats (ark-like, CTM), can group metrics by speaker or other categories, and offers specialized analysis modes like out-of-vocabulary word tracking and oracle WER selection across multiple hypotheses.
The package depends on Levenshtein for edit-distance alignment, numpy for numerical operations, typer for CLI scaffolding, loguru for logging, regex for pattern matching, and termcolor for colored terminal output. It ships prebuilt wheels for Python 3.10, 3.11, and 3.12 across macOS, Linux, and Windows, making installation straightforward on common platforms. Character-aware alignment is available as an option to produce more intuitive word alignments when words are similar, though it can increase reported WER by exposing insertions and deletions that token-only alignment would hide.
Use it for
- Compute aggregate WER/CER metrics on ASR output files to track system performance over time
- Generate detailed per-utterance alignment reports to diagnose specific transcription errors and patterns
- Compare multiple ASR systems or hypotheses against the same reference to rank model quality
- Analyze out-of-vocabulary word recognition rates using OOV-CER for targeted model improvement
- Extract per-speaker or per-group error metrics from mixed-speaker transcriptions for speaker-specific analysis
- Produce JSON summaries of top errors for automated reporting and integration into evaluation pipelines
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes.
texterrors is actively maintained, has no known vulnerabilities, carries a permissive license, and directly solves a common problem in speech recognition evaluation. Medium install friction is acceptable given the prebuilt wheels and straightforward dependency set. Install it if you regularly evaluate ASR or transcription systems.
Install
texterrors on PyPI
Before you install
Medium install friction due to compiled dependencies (Levenshtein, numpy, regex). Active maintenance with recent release (66 days ago). Supports Python 3.9 or newer across multiple platforms via prebuilt wheels.
Requires Python 3.9 or newer; compiled dependencies (Levenshtein, numpy) may require build tools on some platforms.
License in practice
Apache-2.0 permissive license allows use in commercial and proprietary projects with minimal restrictions.
Quickstart
pip install texterrors
from texterrors import align_texts
ref_aligned, hyp_aligned, cost = align_texts(
["speedbird", "eight", "six", "two"],
["hello", "speedbird", "six", "two"],
use_chardiff=True,
)Verify before relying
- Whether the character-aware alignment mode is enabled by default or requires explicit flags
- Performance characteristics on very large transcription files or datasets
- Compatibility with non-English or multilingual transcription outputs
Package facts
| License | Apache-2.0 License permissive |
| Python support | Supports the current Python release >=3.9 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 6 packagesLevenshteinlogurunumpytyperregextermcolor |
| Maintenance | Actively maintained 66 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 875,922 / month, #4,832 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
Evidence: texterrors-1.1.8-cp310-cp310-macosx_10_14_x86_64.whl; texterrors-1.1.8-cp310-cp310-macosx_11_0_arm64.whl; texterrors-1.1.8-cp310-cp310-manylinux_2_17_i686.manylinux2014_i686.whl; texterrors-1.1.8-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; texterrors-1.1.8-cp310-cp310-musllinux_1_2_i686.whl; texterrors-1.1.8-cp310-cp310-musllinux_1_2_x86_64.whl; texterrors-1.1.8-cp310-cp310-win32.whl; texterrors-1.1.8-cp310-cp310-win_amd64.whl; texterrors-1.1.8-cp311-cp311-macosx_10_14_x86_64.whl; texterrors-1.1.8-cp311-cp311-macosx_11_0_arm64.whl; texterrors-1.1.8-cp311-cp311-manylinux_2_17_i686.manylinux2014_i686.whl; texterrors-1.1.8-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; texterrors-1.1.8-cp311-cp311-musllinux_1_2_i686.whl; texterrors-1.1.8-cp311-cp311-musllinux_1_2_x86_64.whl; texterrors-1.1.8-cp311-cp311-win32.whl; texterrors-1.1.8-cp311-cp311-win_amd64.whl; texterrors-1.1.8-cp312-cp312-macosx_10_14_x86_64.whl; texterrors-1.1.8-cp312-cp312-macosx_11_0_arm64.whl; texterrors-1.1.8-cp312-cp312-manylinux_2_17_i686.manylinux2014_i686.whl; texterrors-1.1.8-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “WER CER scoring”
- texterrorsScores automatic speech recognition or transcription output against a…
- jiwerJiWER computes speech recognition evaluation metrics (WER, MER, WIL,…
- nlptuttiMeasures Korean speech-to-text accuracy using character error rate…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also kaldialign · nlptutti · jiwer · rouge · unbabel-comet · pyannote-metrics · rouge-metric · rouge-chinese · whisperx · whisper-timestamped