Carl-bot vs Dyno vs Supervisor: two rule engines and one that reads meaning

Bottom line: pick Carl-bot if you have active moderators, because its punishments combine in one rule and its defer action routes ambiguous cases to a human vote. Pick Dyno if you want the bot to handle it unattended, with 19 filters and an automatic ban ladder. Add Supervisor for the harm neither reads, because both are matching strings and counting events.

Carl-bot and Dyno are the two most moderation-first general purpose bots on Discord, and they are similar enough that choosing between them is genuinely difficult. This post resolves that first, from both bots' own documentation, then covers a third option that does a different job.

The short answer

• Pick Carl-bot if you want the best response system on Discord: composable punishments and a way to route ambiguous cases to human moderators instead of guessing.
• Pick Dyno if you want the most detection surface: nineteen automod filters against Carl-bot's eight, all free.
• Add Supervisor if the messages hurting your community trip neither bot's filters, which is what happens when harm is contextual rather than countable.

At a glance

Carl-bot vs Dyno: the differences that actually matter

Detection surface

Dyno wins on breadth. Nineteen filters: All Caps, Bad Words, Chat Clearing Newlines, Duplicate Text, Character Count, Emoji Spam, Fast Message Spam, Image Spam, Invite Links, Known Phishing Links, Links, Links Cooldown, Mass Mentions, Mentions Cooldown, Spoilers, Masked Links, Stickers, Sticker Cooldown and Zalgo Text.

Carl-bot has eight modules: message spam, attachment spam, bad words, caps limit, honeypot, invites, links and mentions.

The gap is real but narrower than nineteen against eight suggests, because several of Dyno's filters are variations on a theme, and Carl-bot has one Dyno lacks: a honeypot channel, which is a trap for spam bots rather than a filter on human messages. Dyno's genuinely distinct additions are Known Phishing Links, Zalgo Text, Sticker Cooldown and Character Count.

Response model

Carl-bot wins here, decisively. Its punishments are delete, warn, tempmute, mute, timeout, kick, tempban, ban, message, DM and defer. Durations are written naturally as 3h42m, and crucially multiple punishments combine with commas, so delete, tempmute 20m is a single valid response to a rule. Dyno gives you one action per rule from a set of six: Warn, Delete, Auto Mute, Auto Ban, Instant Mute, Instant Ban.

And Carl-bot has defer, which has no equivalent in any bot in this series. Instead of acting automatically, it sends the offending context to a drama channel where moderators decide with reactions. That is the only mechanism here that treats an ambiguous case as something to route to a human rather than something to guess at. It requires premium, and it is the single most thoughtful feature in this category.

Warnings and history

Carl-bot's warns do not automatically expire, and the threshold punishment re-triggers on each new warning while a member is above the limit. That is a deliberate choice that keeps history meaningful.

Dyno's automod warnings are deliberately separate from its manual warning system. Its docs are explicit that the Automod Warn action "is not a ?warn and will not dm the user, be logged in the Modlog channel, appear on ?warnings or ?modlogs, or go toward any Autopunish settings", and that automod violations expire after 5 minutes.

If you want automod incidents to build a lasting record, Carl-bot's model does that and Dyno's does not without extra work.

Pricing model

Both are effectively per server, in different shapes.

Dyno is straightforwardly per server: $5.99 a month or $49.99 a year for Standard, "for one server of your choice". Advanced Automod, meaning per-rule log channels and custom responses, is the paid moderation feature. All nineteen filters and the full ban ladder are free.

Carl-bot sells premium through Patreon from £6.50 a month on a transferable server slot system: you mark servers with /premium addpremium, free slots with /premium removepremium, and list them with /premium listpremium. Tiers grant different numbers of slots. Automod is free; the paid moderation feature is Drama Watcher, which is what powers defer.

Carl-bot's slots being movable without contacting support is a genuine advantage if you spin servers up and down. Dyno's model is simpler to reason about.

Verdict on Carl-bot versus Dyno

Pick Carl-bot if you have active human moderators. The composable punishments and the defer queue are built for a team that wants to make the calls, and the reaction roles are the best on Discord if you need those too.

Pick Dyno if you want the bot to handle it. More filters, finer thresholds, and an automatic ban ladder that works without anyone watching.

Full breakdowns: what Carl-bot costs and what Dyno costs, plus our Carl-bot review and Dyno review.

Where both hit the same ceiling

Both are rule engines, so both share a blind spot that no amount of configuration closes.

Word lists only catch what you wrote down. Carl-bot's censor lists and Dyno's Banned Words fail on the same input: fr33 n1tr0, h8, 5c4m, h a t e, s.c.a.m, homoglyphs, zero-width characters. Dyno's documentation shows the trade-off explicitly, since wildcard mode matching "hi" inside "high" is the cost of catching variations.

Rate limits measure volume, never meaning. Message spam, attachment spam, caps and mentions all ask how many and how fast. A targeted threat or a well-written scam is one calm message that satisfies no condition.

Neither classifies harm by category. No harassment module, no threat module, no scam module, no self-harm module in either. Those are word lists you maintain, per language.

Neither weighs conversation context, and neither reads what is inside an image.

Carl-bot's honeypot catches bots, and Dyno's phishing detection catches known bad links. Neither catches a person being deliberately cruel in correctly spelled English.

What Supervisor adds, and what it does not

Supervisor reads what a message means rather than matching strings and counting events: 16 labels including harassment, hate, threats, scams, self-harm and illegal activity, conversation context and implicit moderation for ambiguous messages, image and video moderation that reads content, and over 100 languages. Billed per account at £13.99 a month, or about £4.99 on the three year cycle, so one subscription covers every server.

What it does not do:

• No bans, kicks or mutes. Delete, Timeout and Warn only, which is a much weaker response set than either bot here. Carl-bot's eleven combinable punishments are in a different league.
• Nothing like defer. No moderator voting queue.
• No counters, no honeypot, no rate limits.
• No reaction roles, embeds, tags, triggers or starboard. Three slash commands.
• No free tier for new accounts, just a seven day trial and £0.25 of moderation on signup.

The pairings in depth: Supervisor vs Carl-bot and Supervisor vs Dyno.

So which should you use?

One bot, moderators active: Carl-bot. The defer queue alone justifies it.

One bot, mostly automated: Dyno. More filters, automatic escalation, all free.

Either bot, and harmful messages still getting through: the other bot will not fix it. Both read strings and count events. Add a layer that reads meaning and keep the rule engine for the countable work, which it does better than any AI would.

Running several servers: Dyno multiplies per server, Carl-bot consumes a slot per server, and Supervisor is one subscription for the account.

The setup we would actually recommend for a busy server is all three layers: Discord AutoMod free and blocking before messages post, Carl-bot or Dyno for countable rules and enforcement, and an AI layer for the messages that read as harmless to a counter.

Why trust this comparison

Everything above about Carl-bot and Dyno comes from their own material: docs.carl.gg for the automod modules, punishments and premium slot system, and Dyno's automod documentation and pricing page. All read on 29 August 2026. On top.gg's 0 to 100 scale Dyno was 87 from 373 ratings that day and Carl-bot 62 from 594.

We make Supervisor, so this is not a neutral comparison and we have not written it as one. What we have done instead is put the other side's advantages in their own section above, in plain terms, and say where we could not read something rather than guessing at it.

On our own side, the figures come from our source and our evaluation set rather than from marketing copy. Supervisor's models are retrained on real moderation feedback, the thumbs up and thumbs down votes people leave on live flags in their own servers, which is the closest thing to customer research this category has. The last full retrain moved average F1 across the 16 labels from 0.794 to 0.941 on our internal evaluation set, with errors on harmless messages down 44 percent, and version 2.2 improved macro-F1 again across all three model tiers. The method and the per-label numbers are in the 2.1 release post.

Where Supervisor leads, and where it does not. On moderation coverage specifically it is the most complete tool in this series: 16 labels where nothing else classifies more than four, over 100 languages, conversation context, and the only one that reads images and video. It is also the narrowest, with no bans, no raid protection and no utility features, which is why every post here recommends running it alongside another bot rather than instead of one.

We have no customer reviews, ratings or testimonials to show you, because we have not collected any. Judge it on your own content in the live demo rather than on our word for it.

For developers and platforms

Carl-bot and Dyno are Discord bots, so neither helps you moderate your own product. Supervisor has a REST API with SDKs for Python, JavaScript, Go, Rust and Java, plus a Platform API for provisioning moderation to your own users. The API reference has the endpoints and rates.

Whichever you pick, test it on your own content. Paste a message that gets past a keyword filter into the live demo and watch it get caught, or add Supervisor to your server.