Every endpoint, scored in the open.

242 of 242 field-level checks passing across 37 endpoints — run against the live production API on 2026-07-08, not a lab build. The same suite gates every prompt and routing change before it ships.

Documents

Document parse
9/9 checks · $0.10/document
100%
Contract analysis
6/6 checks · $0.12/contract
100%
Invoice extraction
10/10 checks · $0.08/invoice
100%
Receipt extraction
6/6 checks · $0.06/receipt
100%
Statement parsing
9/9 checks · $0.12/statement
100%
Resume parsing
7/7 checks · $0.08/resume
100%
Table extraction
4/4 checks · $0.08/document
100%
Document splitting
5/5 checks · $0.10/bundle
100%
Document comparison
3/3 checks · $0.15/comparison
100%

Extract & structure

Field extraction
7/7 checks · $0.04/call
100%
Classification
4/4 checks · $0.02/call
100%
Guaranteed structure
8/8 checks · $0.06/call
100%
Record normalization
9/9 checks · $0.05/batch
100%
Entity matching
9/9 checks · $0.08/run
100%
Batch categorization
10/10 checks · $0.03/batch
100%

Commerce & ops

Chargeback representment
5/5 checks · $0.15/dispute
100%
Collections sequence
11/11 checks · $0.08/sequence
100%
3-way match
6/6 checks · $0.10/match
100%
Quote builder
5/5 checks · $0.08/quote
100%
Fraud triage
9/9 checks · $0.06/check
100%

Communication

Summarize
3/3 checks · $0.03/call
100%
Sentiment
6/6 checks · $0.02/call
100%
Reply drafting
7/7 checks · $0.05/reply
100%
Ticket triage
9/9 checks · $0.03/ticket
100%
Meeting minutes
9/9 checks · $0.06/transcript
100%
Outreach sequence
7/7 checks · $0.06/sequence
100%
Review response
8/8 checks · $0.03/review
100%

Content & moderation

PII redaction
7/7 checks · $0.03/call
100%
Content moderation
6/6 checks · $0.02/item
100%
Brand-voice rewrite
7/7 checks · $0.03/call
100%
Product copy
8/8 checks · $0.05/product
100%

Web & research

Company enrichment
4/4 checks · $0.05/lookup
100%
Research brief
4/4 checks · $0.50/brief
100%
Lead screening
3/3 checks · $0.10/lead
100%

Media & agents

Agent memory
7/7 checks · $0.02/operation
100%
Image generation
2/2 checks · $0.10/image
100%
Text to speech
3/3 checks · $0.05/clip
100%

How this is measured

Each endpoint has a golden suite of realistic inputs with field-level expectations — exact values, dates, totals, enums, and redaction guarantees, never vibes about prose. The runner calls the production API with a real key and scores every expectation individually. Media-only endpoints (transcribe, describe) are verified with generated audio and images through the same production path. The suite re-runs on every prompt, model-routing, or endpoint change; an endpoint that drops below threshold doesn’t ship.