Alternatives · checked October 22, 2026

Different ways to keep teaching within reach.

A short comparison of a church-specific pilot, a general AI assistant, a sermon library, and an in-house build. It compares approaches, not quality or security.

Shepherds LM pilot

One church’s teaching, with proposed pilot terms.

Proposed offer

Starting point

Material it draws on: Each church’s own teachings (current product statement)
How the approach is framed: Proposed: staff and invited congregants use a church-specific assistant

What to confirm

Delivery and ownership: Pilot delivery, security, hosting, data handling, and usage remain unconfirmed
Price or effort: $3,000 once for six weeks is a proposed offer, not a verified delivery fact

General-purpose AI

Example: a ChatGPT custom GPT.

Starting point

Material it draws on: Uploaded knowledge files can inform a custom GPT¹
How the approach is framed: A configurable assistant with files and instructions¹

What to confirm

Delivery and ownership: Check the selected product’s current plan, settings, and terms
Price or effort: Plan-dependent; no like-for-like price comparison made

Church teaching library

Example: a searchable sermon archive.

Starting point

Material it draws on: SermonAudio offers searchable sermons and sermon-grounded AI answers²
How the approach is framed: Browse a library; some services add questions over their archive²

What to confirm

Delivery and ownership: Check whether the library’s scope fits one church’s approved sources
Price or effort: Service terms vary; no like-for-like price comparison made

Build in-house

A system scoped and operated by the church.

Starting point

Material it draws on: Chosen by the church (analysis)
How the approach is framed: Tailored to church requirements (analysis)

What to confirm

Delivery and ownership: The church takes on its chosen build and operating scope (analysis)
Price or effort: No cost or timeline estimate assumed (analysis)

Comparison of Shepherds LM and alternative approaches
Consideration	Shepherds LM pilot Proposed offer One church’s teaching, with proposed pilot terms.	General-purpose AIExample: a ChatGPT custom GPT.	Church teaching libraryExample: a searchable sermon archive.	Build in-houseA system scoped and operated by the church.
Starting point
Material it draws on	Each church’s own teachings (current product statement)	Uploaded knowledge files can inform a custom GPT¹	SermonAudio offers searchable sermons and sermon-grounded AI answers²	Chosen by the church (analysis)
How the approach is framed	Proposed: staff and invited congregants use a church-specific assistant	A configurable assistant with files and instructions¹	Browse a library; some services add questions over their archive²	Tailored to church requirements (analysis)
What to confirm
Delivery and ownership	Pilot delivery, security, hosting, data handling, and usage remain unconfirmed	Check the selected product’s current plan, settings, and terms	Check whether the library’s scope fits one church’s approved sources	The church takes on its chosen build and operating scope (analysis)
Price or effort	$3,000 once for six weeks is a proposed offer, not a verified delivery fact	Plan-dependent; no like-for-like price comparison made	Service terms vary; no like-for-like price comparison made	No cost or timeline estimate assumed (analysis)

Official sources checked

1 · OpenAI Help Center. Its GPT documentation says uploaded files can serve as knowledge sources for a custom GPT. Creating and editing GPTs.
2 · SermonAudio. Its search page describes natural-language questions with answers grounded in sermons available on the service. Search sermons.

The in-house build observations are analysis about ownership and scope, not researched cost or delivery estimates. The dispatch research snapshot was unavailable; this page makes no unsupported claims about comparative accuracy, privacy, security, or quality.

Review the proposed pilot terms.

See the price, limits, delivery questions, and no-charge inquiry form.

Review the pilot offer