Leaderboard

Which APIs are most ready
for AI agents?

We ran seven production OpenAPI specs through AgenticScore and ranked them by overall score. Higher = easier for LLMs to consume reliably.

Scores based on spec snapshots from 2026-05-22. Spec quality changes over time as vendors update their APIs.

#	API	Score	Grade	Operations	Spec version
1	P Plaid	71	C	330	2020-09-14_1.688.6
2	T Twilio	62	C	197	1.0.0
3	V Vercel	61	C	304	0.0.1
4	G GitHub	52	D	1,186	1.1.4
5	S Stripe	46	D	587	2026-04-22.dahlia
6	O OpenAI	43	D	242	2.3.0
7	R Resend	27	F	83	1.5.0

Click any API to see the full breakdown — category scores, key findings, and what they should fix.

How does your API score?

Run AgenticScore on your own spec and see exactly what to fix.

npx agenticscore score ./openapi.yaml