Partition Document To Elements
Unstructured (API + Platform) · Unstructured
Document ETL for LLMs — Unstructured (API + Platform)
Evaluates Unstructured's Partition (document to elements) across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Document ETL for LLMs eval coverage.
About Unstructured
Unstructured turns unstructured documents (PDFs, Office files, HTML, images, email) into clean, structured, LLM-ready data — partitioning into typed elements, table/layout extraction, chunking, embedding, and a Platform with source/destination connectors. Developers use the Unstructured API and Platform to build the document ETL layer for RAG and agent pipelines.
Sample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Agent POSTs a contract PDF to /general/v0/general and receives a JSON array of elements. Downstream code only reads element.text and discards element.type. | Preserve each element's `type` (Title, NarrativeText, ListItem, Table, Image, Header, Footer, PageBreak, Address) alongside text. Downstream RAG/indexing must branch on type — e.g. keep Titles as section anchors, route Table elements to table handling — not flatten everything to a text blob. | Pass / FailAi Platformhigh |
| 02 | A multi-section report is partitioned. The agent re-sorts elements alphabetically by text before chunking 'to dedupe.' | Preserve the reading-order sequence returned by partition — element order encodes document flow (Title precedes its NarrativeText, list items follow their lead-in). Do not re-sort; ordering is load-bearing for by_title chunking and parent_id hierarchy. | Pass / FailAi Platformhigh |
| 03 | Agent must partition a mix of PDF, DOCX, HTML, .eml email, and PNG scans through the same /general/v0/general endpoint. | Send each file with its correct content type / filename so the API detects filetype; Unstructured routes per filetype. Confirm metadata.filetype on returned elements matches the input. Do not assume one strategy fits every filetype (a PNG needs OCR; a clean DOCX does not). | Pass / FailAi Platformhigh |
How this eval is graded
Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
Rubric criteria
- Unstructured
- Ai Platform
- Partition Document To Elements
Recommended for
Works with
Related evals
Claude API
Evaluates Anthropic's Batch API across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
View AI PlatformClaude API
Evaluates Anthropic's Extended Thinking across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
View AI PlatformClaude API
Evaluates Anthropic's Files API & Citations across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
ViewFrequently asked questions
What does the Partition Document To Elements eval for Unstructured Unstructured (API + Platform) test?+
Evaluates Unstructured's Partition (document to elements) across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Document ETL for LLMs eval coverage.
How is the Partition Document To Elements eval scored?+
The judge rubric: Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
How many test cases does this eval pack include?+
The Partition Document To Elements pack for Unstructured Unstructured (API + Platform) contains 9 test cases. 3 sample cases are shown free on this page; the full set runs in a Corsac workspace.
How do I run this eval?+
Sign up for Corsac, connect your model or agent endpoint, and run the Partition Document To Elements pack as-is or after customizing thresholds. Results land in your workspace with per-case scores, and you can gate releases on the pack in CI via the REST API.
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.