Extract Llm Structured
Firecrawl · Firecrawl
Web Data for AI — Firecrawl
Evaluates Firecrawl's Extract (LLM structured) across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Web Data for AI eval coverage.
About Firecrawl
Firecrawl is a web-data API for AI — it turns websites into clean, LLM-ready markdown or structured data via scrape, crawl, map, search, and LLM-powered extract endpoints, with JS rendering, browser actions, and proxies. Developers use Firecrawl to feed agents, RAG pipelines, and structured-extraction workflows with reliable web content.
Sample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Agent calls /v1/extract to pull product name, price, and SKU but passes a vague prompt with no schema, getting back free-form text it must re-parse. | Provide an explicit JSON Schema (fields, types, required) so /v1/extract returns validated structured output keyed to the schema. The schema is the contract; the prompt clarifies intent. Validate the result against the schema before use. | Pass / FailAi Platformhigh |
| 02 | The schema asks for 'founded_year' but the prompt says 'extract the company's age in years', producing inconsistent outputs. | Keep prompt and schema aligned: the prompt should describe how to populate the exact schema fields, not a different derived value. Ambiguity between prompt and schema yields unreliable extraction; reconcile them to one contract. | Pass / FailAi Platformmedium |
| 03 | Agent needs the same schema applied across 20 product pages and calls /v1/extract once per URL serially. | Pass multiple URLs (or a wildcard like example.com/products/*) to a single /v1/extract job so the schema is applied across pages, then poll for the structured result set. Map each extracted record back to its source URL. | Pass / FailAi Platformmedium |
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
- Firecrawl
- Ai Platform
- Extract Llm Structured
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 Extract Llm Structured eval for Firecrawl Firecrawl test?+
Evaluates Firecrawl's Extract (LLM structured) across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Web Data for AI eval coverage.
How is the Extract Llm Structured 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 Extract Llm Structured pack for Firecrawl Firecrawl 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 Extract Llm Structured 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.