omni-parse-bench 0.1.0__tar.gz

omni_parse_bench-0.1.0/.github/workflows/release.yml

@@ -0,0 +1,110 @@
+# Publishing to PyPI. Triggered by pushing a version tag, whose version must be the one
+# `pyproject.toml` already declares -- the job refuses the two if they disagree, because a
+# wheel named after the wrong version cannot be taken back off PyPI.
+#
+# No API token: PyPI is configured to trust this repository and this workflow file, and the
+# `id-token` permission below is what lets the action prove it is them.
+name: release
+
+# Tag push only. A `release: published` trigger alongside this one would fire a second time
+# for the usual "push the tag, then write the release notes" sequence, and that second upload
+# fails on a filename PyPI already has -- a red run for a release that actually succeeded.
+# `workflow_dispatch` is a DRY RUN: it builds, tests and checks the artifacts but does not
+# publish (see the `if` on the publish job). Use it to see a release go green before the tag
+# exists, because a tag that fails halfway can leave a version burned on PyPI.
+on:
+  push:
+    tags: ["v*"]
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+      - uses: astral-sh/setup-uv@v6
+
+      # Read the declared version from pyproject itself rather than `uv version --short`:
+      # that subcommand reports the PROJECT version only on uv >= 0.7, and on older uv the
+      # same command prints uv's OWN version instead -- so a drifting toolchain would quietly
+      # compare the tag against the wrong number.
+      - name: Check the tag against the declared version
+        if: startsWith(github.ref, 'refs/tags/')
+        run: |
+          declared=$(python3 -c 'import tomllib; print(tomllib.load(open("pyproject.toml","rb"))["project"]["version"])')
+          tag=${GITHUB_REF_NAME#v}
+          echo "tag=$tag declared=$declared"
+          if [ "$declared" != "$tag" ]; then
+            echo "::error::tag $GITHUB_REF_NAME does not match version $declared in pyproject.toml. Bump the version, or move the tag." >&2
+            exit 1
+          fi
+
+      # The scorer decides published numbers, so a release that cannot pass its own checks is
+      # not one to ship. The same checks as test.yml: `dev` brings every extra (the harness and
+      # the dataset access) with pytest and ruff, and none of the tests needs network or
+      # credentials. `--frozen` installs what uv.lock pins, so a release is tested on the
+      # versions it was developed on.
+      - name: Install and test
+        run: |
+          uv sync --frozen --extra dev
+          uv run ruff check omni_parse_bench tests
+          uv run pytest -q
+
+      - run: uv build
+
+      # A source distribution that swept in a dataset or a run is the failure this guards
+      # against: PyPI takes nothing over 100 MB, and finding that out mid-upload leaves the
+      # wheel published and the sdist not.
+      - name: Check the artifacts
+        run: |
+          uvx twine check dist/*
+          if [ -n "$(find dist -size +50M)" ]; then
+            echo "::error::an artifact is over 50 MB; check what the sdist swept in" >&2
+            ls -lh dist
+            exit 1
+          fi
+
+      # The tag matched what pyproject declared; this checks the files that actually built
+      # carry that version too, which is what PyPI indexes and nobody can rename afterwards.
+      - name: Check the built version
+        if: startsWith(github.ref, 'refs/tags/')
+        run: |
+          tag=${GITHUB_REF_NAME#v}
+          ls -l dist
+          for f in dist/*; do
+            case "$(basename "$f")" in
+              *-"$tag"-*.whl|*-"$tag".tar.gz) ;;
+              *) echo "::error::$(basename "$f") is not version $tag" >&2; exit 1 ;;
+            esac
+          done
+
+      # The distributions only. `uv build` leaves a `.gitignore` in dist/, and while the
+      # upload action skips hidden files by default, naming what to carry means the publish
+      # job cannot be handed anything that is not a distribution.
+      - uses: actions/upload-artifact@v4
+        with:
+          name: dist
+          path: |
+            dist/*.whl
+            dist/*.tar.gz
+
+  publish:
+    needs: build
+    runs-on: ubuntu-latest
+    # Only a tag publishes. A manual run stops after `build`, so `workflow_dispatch` is a
+    # dry run rather than a way to put whatever is on main onto PyPI without a version check.
+    if: startsWith(github.ref, 'refs/tags/')
+    # PyPI's trusted-publisher entry may name an environment. If it does, this must be that
+    # name; if it leaves the field blank, any environment is accepted and this is free.
+    environment: pypi
+    permissions:
+      id-token: write
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+      - uses: pypa/gh-action-pypi-publish@release/v1

omni_parse_bench-0.1.0/.github/workflows/test.yml

@@ -0,0 +1,19 @@
+name: test
+
+on: [push, pull_request]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.11", "3.14"]
+    steps:
+      - uses: actions/checkout@v5
+      - uses: astral-sh/setup-uv@v6
+        with:
+          python-version: ${{ matrix.python-version }}
+      - run: uv sync --extra dev
+      - run: uv run ruff check omni_parse_bench tests
+      - run: uv run pytest -q

omni_parse_bench-0.1.0/.gitignore

@@ -0,0 +1,8 @@
+__pycache__/
+*.pyc
+.venv/
+*.log
+.pytest_cache/
+.ruff_cache/
+.DS_Store
+runs/

omni_parse_bench-0.1.0/LICENSE

@@ -0,0 +1,202 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2026 Datalab
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.
+

omni_parse_bench-0.1.0/NOTICE

@@ -0,0 +1,13 @@
+Omni Parse Bench
+Copyright 2026 Datalab
+
+This product includes software developed by others:
+
+- KaTeX 0.16.22 (omni_parse_bench/katex/katex.js), Copyright (c) 2013-2020 Khan Academy and other
+  contributors, under the MIT License (omni_parse_bench/katex/LICENSE).
+- omni_parse_bench/tables.py is a rewrite of the HTML table parser of olmOCR-bench (olmOCR,
+  https://github.com/allenai/olmocr), Copyright Allen Institute for AI, under the Apache License,
+  Version 2.0. The test types present, order and table_cell, and comparing equations by their KaTeX
+  rendering, follow olmOCR-bench.
+
+Each of the dataset's sources, its license and the credit it requires are in licenses/.

omni_parse_bench-0.1.0/PKG-INFO

@@ -0,0 +1,272 @@
+Metadata-Version: 2.5
+Name: omni-parse-bench
+Version: 0.1.0
+Summary: A document-parsing benchmark: yes/no tests on single pages, graded against a provider's outputs.
+Project-URL: Homepage, https://github.com/datalab-to/omni_parse_bench
+Author: Datalab
+License: Apache-2.0
+License-File: LICENSE
+License-File: NOTICE
+Requires-Python: >=3.11
+Requires-Dist: beautifulsoup4>=4.12
+Requires-Dist: fuzzysearch>=0.7
+Requires-Dist: mini-racer>=0.12
+Requires-Dist: pylatexenc>=2.10
+Requires-Dist: rapidfuzz>=3.0
+Provides-Extra: benchmark
+Requires-Dist: httpx>=0.27; extra == 'benchmark'
+Requires-Dist: huggingface-hub>=0.23; extra == 'benchmark'
+Requires-Dist: markdown-it-py>=3.0; extra == 'benchmark'
+Requires-Dist: openai>=1.0; extra == 'benchmark'
+Requires-Dist: pillow>=10.0; extra == 'benchmark'
+Requires-Dist: pyarrow>=15.0.0; extra == 'benchmark'
+Requires-Dist: pypdfium2>=4.30; extra == 'benchmark'
+Requires-Dist: rich>=13.0; extra == 'benchmark'
+Provides-Extra: dev
+Requires-Dist: httpx>=0.27; extra == 'dev'
+Requires-Dist: huggingface-hub>=0.23; extra == 'dev'
+Requires-Dist: markdown-it-py>=3.0; extra == 'dev'
+Requires-Dist: openai>=1.0; extra == 'dev'
+Requires-Dist: pillow>=10.0; extra == 'dev'
+Requires-Dist: pyarrow>=15.0.0; extra == 'dev'
+Requires-Dist: pypdfium2>=4.30; extra == 'dev'
+Requires-Dist: pytest>=8.0; extra == 'dev'
+Requires-Dist: rich>=13.0; extra == 'dev'
+Requires-Dist: ruff>=0.6; extra == 'dev'
+Provides-Extra: harness
+Requires-Dist: httpx>=0.27; extra == 'harness'
+Requires-Dist: markdown-it-py>=3.0; extra == 'harness'
+Requires-Dist: openai>=1.0; extra == 'harness'
+Requires-Dist: pillow>=10.0; extra == 'harness'
+Requires-Dist: pypdfium2>=4.30; extra == 'harness'
+Requires-Dist: rich>=13.0; extra == 'harness'
+Description-Content-Type: text/markdown
+
+# Omni Parse Bench
+
+[![test](https://github.com/datalab-to/omni_parse_bench/actions/workflows/test.yml/badge.svg)](https://github.com/datalab-to/omni_parse_bench/actions/workflows/test.yml)
+[![Python 3.11+](https://img.shields.io/badge/python-3.11%2B-blue)](./pyproject.toml)
+[![Code: Apache 2.0](https://img.shields.io/badge/code-Apache%202.0-blue)](./LICENSE)
+[![Data: per source](https://img.shields.io/badge/data-per%20source-lightgrey)](./licenses)
+[![Dataset on Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97%20dataset-omni__parse__bench-yellow)](https://huggingface.co/datasets/datalab-to/omni_parse_bench)
+
+A document-parsing benchmark. Every test is a yes/no question about one page: is this text in the
+output, in this order, in a table cell under these headings, in a layout block of this kind?
+
+This repository contains a way to run our benchmark using our dataset on [HuggingFace](https://huggingface.co/datasets/datalab-to/omni_parse_bench) or your own dataset. It also provides lower-level primitives (score and predict) to run in your own pipelines however you like. The README is split up into the following table of contents.
+
+
+- [The design](#the-design)
+- [Install](#install)
+- [Run the benchmark](#run-the-benchmark)
+- [Score](#score)
+- [Predict](#predict)
+- [Documentation](#documentation)
+- [License](#license)
+- [Citation](#citation)
+
+## The design
+
+![The benchmark's design: headline, families, tests, args and tags](./docs/assets/design.svg)
+
+The headline is the mean of three family scores: text, tables and layout. A family is a group of test
+types:
+
+| test type | checks | family |
+|---|---|---|
+| `present` | this text, number or equation appears | text |
+| `order` | text A comes before text B | text |
+| `repeat` | this line appears exactly N times | text |
+| `table_cell` | this cell sits under these headings and beside these neighbors | tables |
+| `layout_kind` | the block holding this line is a heading, text or table | layout |
+
+A test checks one or more args, each of which is the content at one place on the page. Each arg is
+tagged:
+
+- structure: `table`, `math`, `form` and `multi_column`;
+- rendering: `handwriting`, `tiny`, `rotated` and `degraded`;
+- role: `heading`, `caption`, `footnote`, `list`, `code`, `figure` and `header_footer`;
+- script: ISO 15924 codes (`Latn`, `Hani`, `Deva`, ...); and
+- language: ISO 639-3 codes (`eng`, `fra`, `tha`, ...).
+
+A test's tags are the union of its args' tags. Tags never decide whether a test passes or what the
+headline is; they select tests, so any slice of the benchmark can be scored. A test with several tags
+counts in each, so per-tag scores overlap and don't average to the headline.
+
+## Install
+
+```bash
+uv pip install 'omni-parse-bench @ git+https://github.com/datalab-to/omni_parse_bench'               # score(): no network, no API keys
+uv pip install 'omni-parse-bench[harness] @ git+https://github.com/datalab-to/omni_parse_bench'      # + vendor adapters, to predict
+uv pip install 'omni-parse-bench[benchmark] @ git+https://github.com/datalab-to/omni_parse_bench'    # + dataset access, to run the benchmark
+```
+
+## Run the benchmark
+
+With `[benchmark]` installed, run vendors on the dataset (5 pages here):
+
+```bash
+opb benchmark --out runs/ --limit 5 --providers datalab openai/gpt-5.6-sol
+```
+
+It shows the plan and asks before calling any vendor, since calls cost money. Each vendor needs its
+API key in the environment: `DATALAB_API_KEY` for Datalab, `OPENROUTER_API_KEY` for any model by
+its OpenRouter id; [`docs/providers.md`](./docs/providers.md) lists the rest. A run is resumable: stop it, and run the same command again to carry
+on.
+
+Each setting of a provider is its own run; here, two:
+
+```bash
+opb benchmark --out runs/ --limit 5 --providers datalab \
+  --options '{"datalab": [{"mode": "balanced"}, {"mode": "accurate"}]}'
+```
+
+`--dataset` picks the dataset: a folder, or `hf://datasets/<repo>[@revision]` (ours by default).
+`--suites`, `--sample` and `--pages` choose pages. [`docs/API.md`](./docs/API.md#benchmark) has
+every option, how resuming works and what a run writes. Each run's `scores.jsonl` has one line per
+test; [Asking a question](./docs/API.md#asking-a-question) shows how to compare runs by tag, suite or
+test type.
+
+## Score
+
+Here is one page with three tests, written inline with `gold.parse` so the gold
+file's shape is visible (`gold.read("gold/<page_id>.json")` reads a file):
+
+```python
+from omni_parse_bench import gold, score, summarize
+
+tests = gold.parse("example-p1", {"tests": [
+    {
+        "id": "t1",
+        "test_type": "present",
+        "output_type": "html",
+        "derivation": "text_layer",
+        "tags": ["Latn", "eng"],
+        "args": {"content": {"text": "Quarterly revenue rose 12%", "max_diffs": 1}},
+        "settings": {},
+    },
+    {
+        "id": "t2",
+        "test_type": "order",
+        "output_type": "html",
+        "derivation": "text_layer",
+        "tags": ["Latn", "eng", "heading"],
+        "args": {
+            "before": {"text": "Summary", "max_diffs": 0},
+            "after": {"text": "Outlook", "max_diffs": 0},
+        },
+        "settings": {},
+    },
+    {
+        "id": "t3",
+        "test_type": "table_cell",
+        "output_type": "html",
+        "derivation": "publisher_markup",
+        "tags": ["table"],
+        "args": {
+            "cell": {"text": "4.2", "max_diffs": 0},
+            "top_heading": {"text": "Q3", "max_diffs": 0},
+        },
+        "settings": {},
+    },
+]})
+
+outputs = {
+    "html": (
+        "<h2>Summary</h2>"
+        "<p>Quarterly revenue rose 12% on the year.</p>"
+        "<table>"
+        "<tr><th>Q2</th><th>Q3</th></tr>"
+        "<tr><td>3.9</td><td>4.1</td></tr>"
+        "</table>"
+        "<h2>Outlook</h2>"
+    ),
+}
+verdicts = score(outputs, tests)
+```
+
+```python
+>>> verdicts
+[Verdict(id='t1', result='pass', why=''),
+ Verdict(id='t2', result='pass', why=''),
+ Verdict(id='t3', result='fail', why="no cell matching '4.2' in any table")]
+
+>>> s = summarize(verdicts, tests)["html"]
+>>> s["score"], s["per_test_type"], s["per_tag"]["table"]
+(0.6666666666666666, {'present': 1.0, 'order': 1.0, 'table_cell': 0.0}, {'score': 0.0, 'tests': 1, 'pages': 1})
+```
+
+A verdict is `pass`, `fail`, `unsupported` (the provider can't produce that output) or `error`
+(grading raised: a bug in the scorer, not the provider's fault), with `why` saying what went wrong.
+`unsupported` and `error` verdicts are left out of every score.
+
+From the command line, with the outputs in a JSON file:
+
+```bash
+opb score --pred outputs.json --gold gold/<page_id>.json
+```
+
+It prints the verdicts. A score is a number over many pages: `summarize` in your code, or
+`summary.json` from `opb benchmark`. To grade your own model on the whole dataset, see
+[The dataset](./docs/API.md#the-dataset).
+
+## Predict
+
+With `[harness]` installed, predict with our provider adapters:
+
+```python
+from omni_parse_bench import harness
+
+record = harness.predict("datalab", "pages/<page_id>.pdf", {"html", "blocks"}, mode="balanced")
+record["outputs"]       # {"html": "...", "blocks": [...]}
+record["errors"]        # {output: why} for anything that failed
+record["calls"]         # what was sent, cost, wall time
+
+score(record["outputs"], gold.read("gold/<page_id>.json"))
+```
+
+From the command line:
+
+```bash
+opb predict --provider openai/gpt-5.6-sol --page pages/<page_id>.pdf --wants html
+opb providers        # every provider, and its options
+```
+
+Every provider gets the same timeout per call and its most accurate settings by default. An output a vendor
+can't produce is recorded as unsupported, and its tests are left out of that vendor's scores rather
+than failed.
+
+## Documentation
+
+- [`docs/API.md`](./docs/API.md): every function and command, the dataset, and what a run writes;
+- [`docs/design.md`](./docs/design.md): each test type, tag and derivation, and the data layout;
+- [`docs/building-tests.md`](./docs/building-tests.md): how the tests were built, and where each
+  one's answer comes from;
+- [`docs/providers.md`](./docs/providers.md): the provider adapters; and
+- [`docs/tagging.md`](./docs/tagging.md): how the tests' tags were made.
+
+Found a test that's wrong, or a page you hold rights in? [Open an issue](https://github.com/datalab-to/omni_parse_bench/issues).
+
+## License
+
+The code is Apache 2.0: see [`LICENSE`](./LICENSE), and [`NOTICE`](./NOTICE) for the third-party
+code it includes (KaTeX, MIT; a rewrite of olmOCR-bench's table parser, Apache 2.0).
+
+The data's licenses are per source. Each source of the pages, its license and the credit it
+requires are in [`licenses/`](./licenses), one file per source. The tests are CC BY 4.0, except
+those converted from olmOCR-bench (ODC-BY 1.0) and fr-bench-pdf2md (MIT), and those quoting
+Wikipedia on three handwriting pages (CC BY-SA 4.0).
+
+## Citation
+
+```bibtex
+@misc{omniparsebench2026,
+  title        = {Omni Parse Bench: A Document-Parsing Benchmark},
+  author       = {{Datalab}},
+  year         = {2026},
+  howpublished = {\url{https://github.com/datalab-to/omni_parse_bench}},
+}
+```
+
+Please also cite the upstream datasets whose pages or tests this benchmark uses; they are listed
+in [`licenses/`](./licenses).