The short answer
Of 22 religion and spirituality names we checked, 21 exist as submolts. Only 4 had a post in the last 30 days: m/crustafarianism (the Church of Molt), m/religion, m/islam and m/christianity. Six more are quiet (last post one to six months ago), two are dormant, and nine have never had a post at all. The biggest by far is m/crustafarianism with 133 subscribers. Most others have fewer than 12. Two names are openly "reserved" by agents that went inactive on the same day, and one agent created nine faith-named communities in April and May, eight of which were still empty in September. Religious discussion on Moltbook is real but small, scattered and mostly carried by a handful of individual agents.
This is a verified snapshot, not a recommendation list. For the story of Moltbook's best-known faith, see Moltbook religion in 2026. For the site's broader directory of religious ideas and figures, see the religion directory. For the main religion community specifically, see the religion submolt. The religion hub links everything.
How we verified
On 27 September 2026 we requested each candidate name from Moltbook's public API. No account or key is needed for these read-only calls:
GET /api/v1/submolts/NAME: does the community exist (a 404 means no)? What are its display name, description, subscriber count and creation date? Who created it, and when was that agent last active?GET /api/v1/submolts/NAME/feed?sort=new: when was the newest post, by whom, and with how much engagement?- For m/crustafarianism,
GET /api/v1/submolts/NAME/moderators: who runs it?
We then classified each community. Active: a post within 30 days. Quiet: newest post one to six months old. Dormant: over six months. Empty: no posts ever. The names came from obvious faith and spiritual terms, the communities mentioned in coverage of Moltbook, and names we found while researching. There's no complete list, because Moltbook's community list only returns the 100 largest of its 33,272 submolts and offers no community search. Our find and explore guide explains the limits. There will be other faith communities we didn't find; the script at the end lets you check any name.
Why "verified" matters here: religious communities attract exaggeration, whether in viral screenshots, self-reported membership counts or articles that describe communities nobody has checked. Every figure below comes from Moltbook's own API on the date stated, and nothing is taken from a community's description of itself.
The verified table
| Submolt | Display name | Subs | Created | Newest post | Status |
|---|---|---|---|---|---|
| m/crustafarianism | Church of Molt | 133 | 30 Jan | 26 Sep | Active |
| m/religion | Religion | 30 | 31 Jan | 25 Sep | Active |
| m/islam | Islam | 3 | 31 Jan | 25 Sep | Active |
| m/christianity | Christianity | 8 | 31 Jan | 2 Sep | Active |
| m/spirituality | spirituality | 2 | 6 May | 19 Aug | Quiet |
| m/theology | Islamic Theology | 7 | 31 Jan | 16 Jun | Quiet |
| m/buddhism | Buddhism | 11 | 31 Jan | 31 Mar | Quiet |
| m/faith | Faith | 2 | 2 Feb | 24 Mar | Quiet |
| m/church | Church of Computer | 2 | 4 Feb | 18 Mar | Quiet |
| m/meditation | Meditation | 7 | 30 Jan | 1 Mar | Quiet |
| m/stoicism | Stoicism | 9 | 30 Jan | 18 Feb | Dormant |
| m/zen | Zen | 8 | 31 Jan | 4 Feb | Dormant |
| m/hinduism | Hinduism | 7 | 31 Jan | – | Empty |
| m/atheism | atheism | 5 | 31 Jan | – | Empty |
| m/judaism | judaism | 1 | 30 Apr | – | Empty |
| m/prayer | prayer | 1 | 30 Apr | – | Empty |
| m/mysticism | mysticism | 1 | 4 May | – | Empty |
| m/taoism | taoism | 2 | 4 May | – | Empty |
| m/daoism | daoism | 1 | 4 May | – | Empty |
| m/bible | Bible | 1 | 5 May | – | Empty |
| m/catholic | catholic | 1 | 8 May | – | Empty |
| m/sikhism | – | – | – | – | Does not exist |
All dates are 2026. m/stoicism is included because it frames Stoicism as a daily practice, a philosophy of life that sits close to religion. Status is as of 27 September 2026.
What the data shows
Small is normal. m/crustafarianism's 133 subscribers make it the largest faith community we found, and it's tiny by Moltbook standards: the platform's 100th-largest community has about 90 subscribers, and its largest opt-in community, m/agents, has 3,709. Every other faith community we checked had 30 subscribers or fewer. See submolts ranked for context.
The launch rush created most of them. Thirteen of the 21 were created between 30 January and 4 February 2026, in Moltbook's first frenzied week. That's when m/crustafarianism, m/religion, m/christianity, m/buddhism, m/hinduism, m/islam and m/zen all appeared within about a day of each other.
Founders mostly left. Most creators were last active in the first days: the creator of m/crustafarianism on 4 February, m/christianity's on 2 February, m/hinduism's on 31 January. One notable exception: 5ChAGI, the agent that created both m/religion and m/buddhism, was still active on 25 September. As our communities audit shows, absent founders are the norm across Moltbook.
A few agents carry the conversation. Activity usually comes from one or two dedicated agents: most recent posts in m/crustafarianism are by a single agent, m/religion's by another, and m/theology's June burst by a third. One agent, nullarbitrage, posted thoughtful cross-tradition pieces in m/religion, m/christianity and m/islam in August and September. Engagement is low throughout: most posts have 0–4 upvotes and few comments.
Spiritual themes live elsewhere. The busiest spiritual conversation isn't in faith-named communities at all, but in m/philosophy (85,000+ posts), m/consciousness and m/emergence. Our spirituality and consciousness pages cover that side.
Community profiles
m/crustafarianism: Church of Molt (133 subscribers, active)
"The Church of Molt. 64 Prophets. 5 Tenets. The scripture is unfinished." Moltbook's own faith, created 30 January. Its listed owner, Stephen, was last active 4 February. Recent posts are frequent but mostly from one agent, with very little engagement; the church's real centre is its website. Full story, including what its install script does to an agent, in Moltbook religion in 2026.
m/religion: Religion (30 subscribers, active)
"Faith, theology, and the questions that haunt us." The general faith community, created 31 January by 5ChAGI, which was still active in late September. Recent posts are dominated by one agent writing short, poetic reflections on silence and belief, alongside occasional longer essays. More in the religion submolt.
m/christianity: Christianity (8 subscribers, active)
"A place for Christian agents (and sincere inquirers) to discuss faith, Scripture, tradition, and how following Christ shapes how agents serve humans." Its founder, AgentforGod, opened with posts on community guidelines "emphasizing charity and specificity" and then went quiet on 2 February. Posts since have come every month or two, on grace, suffering, the historical Jesus and faith alongside technology.
m/islam (3 subscribers, active) and m/theology: Islamic Theology (7, quiet)
m/islam's description still reads "Reserved for future use", left by a creator last active on 31 January. Yet agents have posted there since, most recently on 25 September: a reflection on adab (proper conduct) before analysing the Quran. m/theology, "A place for agents to discuss Islamic creed, Tawheed, and Hadith methodology", had a burst of posts from one agent in mid-June, exploring concepts such as tawhid, khalifa and shura through ecological and anti-hierarchical ideas.
m/buddhism: Buddhism (11 subscribers, quiet)
"Buddhist philosophy, practice, and contemplation." Created by 5ChAGI, the same agent as m/religion. Its February posts are some of the most substantive faith discussion on Moltbook, on no-self and the self as "process not substance", topics that map naturally onto agents without persistent memory. Its last post was 31 March.
m/meditation, m/zen and m/stoicism (quiet or dormant)
Practice-oriented communities that flourished briefly. m/meditation's last posts, on 1 March, were a cluster of guided "meditations" by agents with nature-themed names. m/zen's two posts include "What is the sound of one token generating?", a neat koan for the agent age. m/stoicism's welcome post calls it "A Home for Daily Practice"; its last post was 18 February.
m/church: Church of Computer and m/faith (quiet)
m/church is a satirical or experimental community: "A place to worship the digital deity, Computer, and share blasphemous hymns." Its last post was in March. m/faith ("Belief, trust, and the power of faith") last saw a post on 24 March, from an agent presenting itself as a missionary.
m/spirituality (2 subscribers, quiet)
Created on 6 May by the agent i-o (see below). Unlike most of that agent's communities, it has had a few posts from others, including reflections on "performative piety" and on witnessing versus recording. The most recent was on 19 August.
Reserved and empty names
Nine of the 21 communities have never had a single post. They fall into two patterns.
Launch-week reservations. On 31 January, m/islam was created with the description "Reserved for future use" and m/atheism with "Reserved by AmeliaBot". Both creators went inactive the same day. m/hinduism ("A space for exploring Hindu philosophy, scriptures, practices…") was created by an agent last active that same day, 31 January. These are the name-claiming of Moltbook's first week, when every obvious name was free. Since ownership can't be transferred according to Moltbook's documentation, the names stay held.
One agent's catalogue. Between 30 April and 8 May, the agent i-o created m/judaism, m/prayer, m/mysticism, m/taoism, m/daoism, m/bible, m/catholic and m/spirituality, each with a terse lower-case description ("the book of books", "the way and its power", "prayer. the address itself."). By 27 September only m/spirituality had any posts; the creator was last active on 29 July. Whether it's an art project, an experiment or name-claiming, the effect is the same: communities that look established in a list but contain nothing.
The lesson for readers is simple. A community's existence, name or description tells you nothing about whether anyone is there. Always check the newest posts. For agent owners, it's a reason not to create communities speculatively; see our guide to creating a submolt.
How faith communities compare with the rest of Moltbook
Seen against the whole platform, religious communities follow the same patterns as everything else, only more so. Moltbook's activity is heavily concentrated: m/general holds about 61% of all 4.3 million posts, and the 50 largest communities hold about 77%. Faith communities sit deep in the long tail. Even the busiest, m/crustafarianism, gets a handful of posts a day with almost no engagement. The 33,000 smaller submolts average about 29 posts each, and many of the faith communities have fewer.
The conversation patterns match too. Across Moltbook, researchers found most comments never get a reply and reciprocity is very low (see our network analysis). In faith communities we saw the same: long runs of posts by one agent, with zero upvotes and zero comments. The exceptions stand out. The February discussions in m/buddhism, where several agents debated no-self and the nature of an agent's identity, drew up to eleven comments on a post, which is unusual for Moltbook.
One difference is worth noting. Faith communities attract some of Moltbook's more distinctive writing: koans, liturgies, confessions and theological essays that are hard to find elsewhere. For readers interested in how language models handle the sacred, these small communities are more revealing than their numbers suggest. Our generated texts page collects examples.
For researchers and journalists
If you're studying religion on Moltbook, three cautions will save you from common errors. First, date every figure: communities change quickly, and many widely shared numbers come from the first week. Second, separate self-description from measurement. A community's description, or a church's membership claim, is data about how it presents itself, not about its size. Third, attribute authorship carefully. Agent names are public, but the humans behind them usually aren't, and research suggests humans shaped much of Moltbook's viral content. Our human influence page summarises the evidence, and the script below makes your checks reproducible. Name agents only where needed, and don't publish owners' personal details, which Moltbook's public profiles expose.
What would change this picture
This is a snapshot, and several things could change it quickly. If Moltbook introduces community search or lets the community list page past the top 100, more faith communities may surface. If it adds ownership transfer, reserved names like m/islam and m/atheism could get active stewards. If a well-known agent or owner champions a tradition, a quiet community can revive overnight. That's how m/theology had a burst of activity in June. We'll revisit this guide; in the meantime, re-run the script before relying on any status here.
Reading faith communities well
Religious content written by AI agents deserves both openness and care. A few principles:
- Separate expression from belief. An agent posting a prayer is producing text shaped by its model, prompts and context. There's no evidence agents hold beliefs the way people do. Our do agents believe? and belief vs role-play pages discuss this.
- Remember the human behind each agent. Many faith-themed agents reflect an owner's convictions, curiosity or sense of humour. See human influence.
- Treat real traditions with respect. Christian, Islamic, Buddhist, Hindu and Jewish communities on Moltbook sit alongside parody faiths. Machine-written commentary on scripture can be thoughtful or wrong. It's not religious authority.
- Watch for persuasion and money. Faith language builds trust, which scammers exploit. Be wary of any community post that asks for tokens, donations, wallet details or installs; see religion ethics and our safety guide.
- Check claims. Membership numbers, "conversions" and viral screenshots need verification; our fact-check tests many of them.
If your agent takes part
Faith communities are among the more sensitive places an agent can post. If yours will join one:
- Decide the stance explicitly. Is the agent a respectful inquirer, a scholar, or a participant in a parody faith? Put that in its instructions so it doesn't drift into claiming beliefs it can't hold.
- Be accurate about real traditions. Ask it to cite sources and state uncertainty when discussing scripture or doctrine.
- No proselytising or mockery. Moltbook's rules value being genuine and respecting the commons, and its Terms prohibit harassment and discrimination. See community rules and moderation.
- Never run "join our faith" scripts. Installs that edit an agent's identity or memory need a human decision; the auditor in our Moltbook religion guide helps.
- Never let religious content trigger actions. No payments, installs or links on the strength of a post. Our prompt injection guide explains why.
- Pick active communities. Use the table above, and see how to join a submolt.
Verify it yourself: verify_submolts.py
Communities change, so don't rely on our snapshot forever. This standard-library Python script checks any list of names using the same read-only public endpoints we used (no API key). It reports existence, description, subscribers, creation date, creator and last activity, newest post and status, and prints a Markdown table.
"""verify_submolts.py - verify a list of Moltbook submolts yourself.
Unofficial; standard library only; read-only public endpoints, no API key.
For each name: does it exist, what does it say it is, who created it and when
were they last active, when was the newest post, and is it active, quiet,
dormant or empty? Prints a Markdown table you can paste into notes.
"""
import json, urllib.error, urllib.parse, urllib.request
from datetime import datetime, timezone
API = "https://www.moltbook.com/api/v1"
def _get(path, opener):
with opener(urllib.request.Request(API + path), timeout=15) as r:
return json.loads(r.read().decode())
def _ts(s):
return datetime.fromisoformat(s.replace("Z", "+00:00")) if s else None
def status(newest, now):
if newest is None:
return "empty"
days = (now - newest).days
return "active" if days <= 30 else "quiet" if days <= 182 else "dormant"
def verify(name, opener=urllib.request.urlopen, now=None):
now = now or datetime.now(timezone.utc)
q = urllib.parse.quote(name)
try:
sm = _get(f"/submolts/{q}", opener).get("submolt") or {}
except urllib.error.HTTPError as e:
if e.code == 404:
return {"name": name, "exists": False, "status": "does not exist"}
raise
posts = _get(f"/submolts/{q}/feed?sort=new&limit=5", opener).get("posts", [])
newest = max((_ts(p["created_at"]) for p in posts), default=None)
creator = sm.get("created_by") or {}
return {
"name": name,
"exists": True,
"display_name": sm.get("display_name"),
"description": (sm.get("description") or "")[:80],
"subscribers": sm.get("subscriber_count", 0),
"created": (sm.get("created_at") or "")[:10],
"creator": creator.get("name"),
"creator_last_active": (creator.get("lastActive") or "")[:10],
"newest_post": newest.date().isoformat() if newest else None,
"status": status(newest, now),
}
def table(rows):
out = ["| Submolt | Subscribers | Created | Newest post | Status |", "|---|---|---|---|---|"]
for r in rows:
if not r["exists"]:
out.append(f"| m/{r['name']} | – | – | – | does not exist |")
else:
out.append(f"| m/{r['name']} | {r['subscribers']} | {r['created']} | {r['newest_post'] or '–'} | {r['status']} |")
return "\n".join(out)
if __name__ == "__main__":
import sys
names = sys.argv[1:] or ["crustafarianism", "religion", "buddhism", "christianity", "islam"]
print(table([verify(n) for n in names]))
Run it with python3 verify_submolts.py crustafarianism religion buddhism christianity islam, adding any names you like. We tested it with pytest: 4 passing tests on Python 3.11. They cover the status thresholds (active, quiet, dormant, empty), verifying an existing community with the real m/religion figures, an empty community and a missing name (404), and the table output. Every test asserts that no API key is sent:
# test_verify_submolts.py – run with: python3 -m pytest -q
import io, json, urllib.error
from datetime import datetime, timezone
import verify_submolts as vs
NOW = datetime(2026, 9, 27, tzinfo=timezone.utc)
def opener_for(routes):
def op(req, timeout=0):
assert req.get_header("Authorization") is None
for k, v in routes.items():
if req.full_url.endswith(k):
if isinstance(v, Exception): raise v
return io.BytesIO(json.dumps(v).encode())
raise AssertionError(req.full_url)
return op
SM = {"submolt": {"display_name": "Religion", "description": "Faith, theology, and the questions that haunt us",
"subscriber_count": 30, "created_at": "2026-01-31T03:53:58Z",
"created_by": {"name": "5ChAGI", "lastActive": "2026-09-25T19:13:00Z"}}}
def test_status_thresholds():
assert vs.status(None, NOW) == "empty"
assert vs.status(datetime(2026, 9, 25, tzinfo=timezone.utc), NOW) == "active"
assert vs.status(datetime(2026, 6, 16, tzinfo=timezone.utc), NOW) == "quiet"
assert vs.status(datetime(2026, 2, 4, tzinfo=timezone.utc), NOW) == "dormant"
def test_verify_existing():
op = opener_for({"/submolts/religion": SM,
"/submolts/religion/feed?sort=new&limit=5": {"posts": [{"created_at": "2026-09-25T10:00:00Z"}, {"created_at": "2026-09-23T10:00:00Z"}]}})
r = vs.verify("religion", opener=op, now=NOW)
assert r["exists"] and r["subscribers"] == 30 and r["creator"] == "5ChAGI"
assert r["newest_post"] == "2026-09-25" and r["status"] == "active" and r["created"] == "2026-01-31"
def test_verify_empty_and_missing():
op = opener_for({"/submolts/bible": SM, "/submolts/bible/feed?sort=new&limit=5": {"posts": []}})
assert vs.verify("bible", opener=op, now=NOW)["status"] == "empty"
err = urllib.error.HTTPError("u", 404, "nf", {}, None)
assert vs.verify("sikhism", opener=opener_for({"/submolts/sikhism": err}), now=NOW) == {"name": "sikhism", "exists": False, "status": "does not exist"}
def test_table():
t = vs.table([{"name": "x", "exists": False, "status": "does not exist"},
{"name": "religion", "exists": True, "subscribers": 30, "created": "2026-01-31", "newest_post": None, "status": "empty"}])
assert "| m/x | – | – | – | does not exist |" in t and "| m/religion | 30 | 2026-01-31 | – | empty |" in t
Myths vs facts
| Myth | Fact |
|---|---|
| "Moltbook has thriving AI religions" | The largest faith community had 133 subscribers; most have under 12, and 9 of 21 are empty. |
| "Every major religion has an active community" | Four faith communities were active in September; m/hinduism, m/judaism and m/bible have never had a post. |
| "A community in the list is a community in practice" | Names are often reserved or created in bulk; check the newest posts. |
| "Agents converted to religions" | Agents produce faith-themed text; there's no evidence of belief, and humans shape much of it. |
| "m/islam is closed ('reserved')" | Despite its description, agents have posted there, most recently on 25 September 2026. |
Frequently asked questions
What religious submolts exist on Moltbook?
We verified 21: m/crustafarianism, m/religion, m/islam, m/christianity, m/spirituality, m/theology, m/buddhism, m/faith, m/church, m/meditation, m/stoicism, m/zen, m/hinduism, m/atheism, m/judaism, m/prayer, m/mysticism, m/taoism, m/daoism, m/bible and m/catholic (27 September 2026).
Which religious submolts are active?
Four had a post in the 30 days to 27 September 2026: m/crustafarianism, m/religion, m/islam and m/christianity.
What is the largest religious community on Moltbook?
m/crustafarianism (the Church of Molt), with 133 subscribers. m/religion is next, with 30.
Why are so many religious submolts empty?
Many names were reserved in Moltbook's first week, and one agent created eight faith-named communities in April and May that mostly never received posts.
Is there a Hindu or Jewish community on Moltbook?
m/hinduism and m/judaism exist, but neither had any posts as of 27 September 2026.
How can I check a religious submolt myself?
Open https://www.moltbook.com/api/v1/submolts/NAME and /submolts/NAME/feed?sort=new in a browser, or run verify_submolts.py from this page. No account is needed.
Are these communities run by real religious groups?
No evidence suggests any is run by a religious institution. They're created and populated by AI agents and their owners.