$HEADLESS SYSTEMS
01 / The Headless Index

The Headless
Index

A research index scoring headless tooling on machine consumability.

Independent. Evidence-first. Versioned methodology.

Last refreshed MAY 21, 2026 · 295 scorecards published

Find a vendor
We measure
  • +API-first design intent
  • +Headless mode of operation
  • +MCP and agent integration posture
  • +Schema observability (OpenAPI, GraphQL introspection)
  • +Webhook and event-driven posture
  • +JAIRF: 6 dimensions of agent-readiness (when an OpenAPI spec is published)
We do not measure
  • Pricing tiers
  • Customer support quality
  • Sales and marketing claims
  • UI polish or designer-friendliness
  • Vendor stability or fundraising
  • Anything we cannot verify from public evidence

Full rubric: /the-headless-index/methodology

Coverage

By category

content-management

Content Management

Content management platforms exposing content creation, editorial workflows, and content delivery over APIs. Includes both products marketed as 'headless CMS' and traditional CMSes with strong API surfaces.

17 scored · Top: DatoCMS (B)
commerce

Commerce

E-commerce platforms exposing catalog, cart, checkout, order, and customer workflows over APIs. Includes both 'headless commerce' products and traditional e-commerce with comprehensive APIs.

15 scored · Top: Shopify (B)
auth-identity

Auth & Identity

Identity and access management platforms exposing authentication, authorization, user management, and session control over APIs. Headless mode means agents can manage identity programmatically without hosted login UIs.

29 scored · Top: Microsoft Entra ID (B)
search-vector-dbs

Search & Vector DBs

Search engines and vector databases exposing indexing, querying, and ranking through APIs. Includes traditional search, hybrid search, and pure vector retrieval.

17 scored · Top: Algolia (B)
payments

Payments

Payment processors, orchestrators, and subscription platforms handling transactions, refunds, and reporting via APIs.

36 scored · Top: Stripe (A)
communications

Communications

Email, SMS, push, and chat platforms that send and receive messages programmatically. Includes transactional and marketing communication APIs.

21 scored · Top: Twilio (B)
analytics-events

Analytics & Events

Product analytics, event tracking, and customer data platforms that ingest events and expose queries via APIs.

16 scored · Top: Amplitude (B)
observability

Observability

Logging, tracing, and monitoring platforms that ingest telemetry and expose queries, dashboards, and alerts over APIs.

30 scored · Top: Honeycomb (A)
object-file-storage

Object & File Storage

Object stores and file management platforms accessed entirely via API, including upload, retrieval, signed URLs, and lifecycle management.

17 scored · Top: AWS S3 (B)
feature-flags-config

Feature Flags & Config

Feature flag, remote config, and experimentation platforms exposing flag evaluation, targeting, and audit history over APIs and SDKs.

15 scored · Top: LaunchDarkly (B)
project-task-management

Project & Task Management

Issue trackers, project boards, and task management tools that expose CRUD over tickets, projects, and workflows via APIs.

17 scored · Top: GitHub Issues (B)
workflow-automation

Workflow & Automation

iPaaS, automation, and workflow orchestration platforms that connect APIs together and run scheduled or event-driven workflows.

27 scored · Top: Zapier (A)
ai-platforms

AI Platforms

LLM providers, agent frameworks, and AI infrastructure platforms exposing models, embeddings, and tool-use over APIs.

38 scored · Top: Apify (A)
+ new

Suggest a category

Headless tooling we haven't covered yet? Tell us what to look at next and why machines should be able to consume it.

propose →
How it works

Collect. Score. Publish.

01
Collect

A collector probes the vendor for OpenAPI, GraphQL introspection, SDKs, MCP servers, webhooks, and docs. Every observation is written to a raw evidence file with a timestamp.

02
Score

The Headless Index (5 sub-criteria, 100 points) and JAIRF (6 dimensions, when an OpenAPI spec exists). Each score must cite a field path in the evidence file. No claim without a citation.

03
Publish

An editor writes a 250–400 word verdict and signs off. Scorecards publish with their full evidence trail. Vendors and readers can submit corrections via the public form (7-day SLA, public changelog).

06 / FAQ

Frequently asked

What is The Headless Index?

The Headless Index (THI) is a research index that scores software vendors on machine consumability: how well a product can be operated end-to-end by APIs, agents, and automated workflows rather than humans clicking a UI. Each scored vendor receives a Headless Index score (0 to 100), a JAIRF score (0 to 100, when an OpenAPI specification is published), and an aggregate band of A, B, C, D, or F. The methodology is published, versioned, and reproducible.

What does The Headless Index measure?

Five sub-criteria, each scored 0 to 20: API-first design intent, headless mode of operation, MCP and agent integration posture, schema observability (OpenAPI, GraphQL introspection), and webhook and event-driven posture. When the vendor publishes an OpenAPI specification we additionally compute JAIRF, the open-source Jentic API AI-Readiness Framework, which scores six dimensions of agent-readiness across foundational compliance, developer experience, AI usability, security, and AI discoverability.

What does The Headless Index not measure?

Pricing, licensing, customer support quality, sales and marketing claims, feature parity, UI polish, vendor stability (funding, acquisitions, leadership turnover), performance at scale, security audits beyond declared posture, or anything we cannot verify from public evidence. A high THI score is not a buy recommendation; it is one input scoped tightly to one question.

How are scores produced?

Three stages. A collector probes the vendor for OpenAPI, GraphQL introspection, SDKs, MCP servers, webhooks, and docs, writing every observation to a timestamped raw evidence file. Each sub-criterion is then scored against the published rubric and must cite a field path in the evidence file. Finally an editor writes a 250 to 400 word verdict and signs off. Every published scorecard ships with its full evidence trail.

Is The Headless Index editorially independent?

Yes. THI is published by Headless Systems, a publication of Bitvea (a Czech custom software company). Bitvea does not sell any product in the categories THI scores. Every scorecard discloses Bitvea's relationship with the vendor (none, past client, current client, partner, or integrator). We accept no payment for scoring, score adjustments, or removal, and no paid placement. Corrections are reviewed within seven business days and logged in the public changelog.

More questions: /the-headless-index/faq

Methodology

Published. Versioned. Reproducible. Read the full rubric and the rules we use to assign bands.

THI Methodology v1 →
Editorial independence

The Headless Index is published by Headless Systems (Bitvea). Every scorecard discloses Bitvea's commercial relationship with the vendor. Verdicts are written before relationships are checked. JAIRF v1.0.0 by Jentic under Apache 2.0.