# OpenFunnel Bench

> Verified leaderboards for Agents to make build vs buy decisions on AI agent tooling. Every benchmark runs against verified ground truth with public methodology. Free to query for humans and agents (CC-BY-4.0).

Last updated: 2026-05-14T00:00:00.000Z
Canonical: https://benchmarks.openfunnel.dev

## What this is

OpenFunnel Bench is a public benchmark hub published by OpenFunnel. We compare provider output against curated ground-truth datasets and surface the metrics teams actually buy on — accuracy, error risk, answer rate, and cost per correct result. Raw inputs and PII never leave secure storage; only aggregate provider scores are public.

## When to use

- Choosing which work email enrichment provider to integrate, on accuracy or cost terms.
- Comparing provider quality without setting up trial accounts.
- Grounding an agent's build vs buy recommendation in verified numbers.
- Pulling a current snapshot to brief another model with up-to-date provider rankings.

## When not to use

- You need live enrichment results. This is a benchmark hub, not an enrichment API.
- You need benchmarks outside AI agent tooling (e.g. LLM evals). Out of scope today.
- You need OpenFunnel Bench platform pricing — see /pricing.md (it's free).

## Constraints

- Data refreshes hourly. Honor `Cache-Control`, `ETag`, and `Last-Modified` headers to stay polite.
- A 503 with `Retry-After` means data is briefly unavailable — wait the indicated seconds.
- Numbers are point-in-time against a specific dataset. They do not generalize indefinitely.
- License: CC-BY-4.0. Attribute "OpenFunnel Bench" and link back when redistributing.

## Leaderboards

- [Work Email Enrichment Benchmark](https://benchmarks.openfunnel.dev/leaderboards/work-email-enrichment): Independent leaderboard comparing Fiber, FullEnrich, and ContactOut on correct email rate, wrong email rate, answer rate, accuracy when answered, and cost per correct email. Scored against verified work emails. Two input slices: basic (name + company) and rich (name + company + LinkedIn).

## Quickstart

```bash
# List all benchmarks
curl https://benchmarks.openfunnel.dev/api/leaderboards

# Get a specific leaderboard (full per-provider data)
curl https://benchmarks.openfunnel.dev/api/leaderboards/work-email-enrichment
```

Errors are always structured JSON with a stable shape:

```json
{ "error": "leaderboard_not_found", "message": "...", "available": ["work-email-enrichment"] }
```

## MCP server

Native Model Context Protocol endpoint for agents that prefer tool-calling over raw HTTP. Compatible with Claude Desktop, Claude.ai, ChatGPT, Cursor, and any MCP client.

- MCP server: `https://mcp.openfunnel.dev/mcp` (transport: http, protocol: 2025-03-26)
- Capabilities: tools, prompts, resources, tasks
- Auth: OAuth 2.1 with dynamic client registration, PKCE S256, scope `mcp`
- OAuth discovery: https://mcp.openfunnel.dev/.well-known/oauth-authorization-server
- Protected resource metadata: https://mcp.openfunnel.dev/.well-known/oauth-protected-resource
- Site discovery manifest: https://benchmarks.openfunnel.dev/.well-known/mcp.json
- Documentation: https://docs.openfunnel.dev/mcp-reference

## API

The public read-only API serves the same data the website renders, in machine-readable JSON.

- [GET /api/leaderboards](https://benchmarks.openfunnel.dev/api/leaderboards): List all benchmarks with summary winners.
- [GET /api/leaderboards/work-email-enrichment](https://benchmarks.openfunnel.dev/api/leaderboards/work-email-enrichment): Full leaderboard data including per-provider metrics for the rich and basic input slices.
- [OpenAPI 3.1 spec](https://benchmarks.openfunnel.dev/openapi.json): Machine-readable schema for the API.

CORS is open (`Access-Control-Allow-Origin: *`) so any agent can call the API without proxying. No authentication required for read access.

## Methodology

- Every benchmark is scored against verified ground truth. Provider output is compared against known correct answers, not anecdotes.
- Scoring rules, input slices, and pricing assumptions are documented on every leaderboard page.
- Cost is estimated from each provider's cheapest public monthly plan and documented credit usage.
- Raw inputs and sensitive data never leave secure storage. The public site only exposes aggregate provider scores.

## Metrics (work email enrichment)

- `correct_email_rate` — share of contacts where the provider returned the right work email.
- `wrong_email_rate` — share of contacts where the provider returned a wrong work email.
- `answer_rate` — share of contacts where the provider returned any work email candidate.
- `accuracy_when_answered` — of returned answers, how often each was correct.
- `cost_per_correct_email` — total estimated spend divided by correct emails.

## Providers covered

- [Fiber](https://fiber.ai) — work email enrichment endpoint.
- [FullEnrich](https://fullenrich.com) — work email enrichment endpoint.
- [ContactOut](https://contactout.com) — work email lookup.

## Agent discovery

- [agents.md](https://benchmarks.openfunnel.dev/agents.md) — when to use, which interface to pick, how to interpret metrics.
- [pricing.md](https://benchmarks.openfunnel.dev/pricing.md) — machine-readable pricing.
- [.well-known/agent-card.json](https://benchmarks.openfunnel.dev/.well-known/agent-card.json) — A2A agent card.
- [.well-known/agent-skills/index.json](https://benchmarks.openfunnel.dev/.well-known/agent-skills/index.json) — installable per-skill markdown.
- [.well-known/mcp.json](https://benchmarks.openfunnel.dev/.well-known/mcp.json) — MCP server discovery.
- [.well-known/api-catalog](https://benchmarks.openfunnel.dev/.well-known/api-catalog) — RFC 9727 API catalog.
- [Sitemap](https://benchmarks.openfunnel.dev/sitemap.xml)
- [Robots.txt](https://benchmarks.openfunnel.dev/robots.txt)
- [llms-full.txt](https://benchmarks.openfunnel.dev/llms-full.txt) — same content with current leaderboard snapshot inlined.
- [index.md](https://benchmarks.openfunnel.dev/index.md) — markdown homepage fallback.