Free for every server
One product. Two limits.
Every server gets the full scripting engine, every trigger, and the complete API. Base covers normal use; Founder is a thank-you for the servers that backed Mallard early.
See every limit ↓Base
Everything Mallard does, for every server.
- 500s shared runtime/hr
- 10 MB storage
- 60 fetch() calls/hr
- 8 YouTube channels
Founder
For the servers that backed Mallard early. Assigned by hand.
- 1000s shared runtime/hr
- 100 MB storage
- 100 fetch() calls/hr
- 12 YouTube channels
How budgets work
One budget, shared by your scripts.
Runtime is the time your server's scripts spend running, including time waiting for Discord and other network calls. It's measured over a rolling hour.
What happens at the limit?
When the runtime budget is exhausted, new runs are throttled until usage falls within the rolling window. Earlier actions aren't rolled back. Fetch calls and storage have their own limits.
Read how limits work →Compare Base and Founder
Runtime and fetch are hourly budgets measured on a rolling window. Storage is persistent across your key-value store and media uploads. YouTube channels and RSS/Atom feeds have separate limits.
| Limit | Base | Founder |
|---|---|---|
Triggers included
Slash commands, reactions, member joins, Twitch stream events, and YouTube video/stream releases. | All 84 triggers | All 84 triggers |
Script editor
Real-time autocompletion against the full script API and syntax highlighting. | Monaco, full API | Monaco, full API |
OpenRouter AI & integrations
Connect OpenRouter AI, Twitch, YouTube, and signed webhooks on the Integrations page without exposing credentials to scripts or moderators. | Included | Included |
Execution logs & audit
Every script run is recorded with status, duration, error details, and an audit trail. | Included | Included |
Runtime / hour
How long your scripts may spend running each hour, everything combined. Time spent waiting for Discord to answer counts, so this is also what limits how many messages, bans and role changes your scripts can make: there is no separate cap on those. | 500s | 1000s |
Storage (KV + Media)
Persistent storage shared between your scripts' key-value store (settings, moderation history, custom data) and uploaded media files. The one limit here that does not reset every hour, so freeing space means deleting keys or unused media. | 10 MB | 100 MB |
fetch() calls / hour
Outbound HTTP requests your scripts make to other sites and APIs. A separate limiter also caps bursts at 30 requests a minute per server. | 60 | 100 |
Concurrent scripts
How many of this guild's scripts may run at the same time. A slow script no longer blocks quick ones behind it; incrKV/updateKV lock a key so two of them can't race on the same value when they overlap. | 4 | 8 |
YouTube channels
How many YouTube channels this server may link on the Integrations page to receive video and live stream triggers. | 8 | 12 |
RSS / Atom feeds
How many feeds (blogs, subreddits, GitHub releases, status pages) this server may follow on the Integrations page, and how often they're checked for new items. | 20, every 15 min | 50, every 5 min |
Incoming webhooks
Scripts with their own URL that GitHub, Ko-fi, Stripe, uptime monitors or your own services can call, and how many requests each accepts an hour. | 20 URLs, 1,000 req/h each | 50 URLs, 3,000 req/h each |
Leaderboard size
How many members one leaderboard set (zIncr/zRange) can rank. Each member also counts towards storage. | 100,000 members | 100,000 members |
Pending delayed jobs
How many runLater(), emit() and expectReply() jobs a server may hold at once. | 1,000 | 1,000 |
Bot profile from scripts
setBotProfile() lets scripts change Mallard's nickname, avatar, banner and bio in this server, e.g. seasonal avatars on a schedule. | Included | Included |
The same on every server
Safety limits that protect the runner and the database. They are not sold as extras, and they do not change between Base and Founder.
| Rule | Value |
|---|---|
Script size
Checked when a script is saved. |
60,000 characters |
Scheduled interval
Enforced at save time and again at dispatch time; cron schedules may not fire more often either. |
At least 5 minutes |
Script run timeout
A single run that takes longer is stopped by the runner. |
40 seconds |
runLater() delay
Longest a script may schedule a future run. |
90 days |
Job payload
Largest runLater(), emit() or expectReply() payload, serialized. |
4 KB |
expectReply() wait
Longest a script may wait for a reply. |
24 hours |
fetch() burst
On top of the hourly budget, and separate from Discord API calls. |
30 requests/minute per server |
Discord API
Mallard does not add a separate cap on Discord actions; their wait time is counted in your runtime. |
Discord's own limits |
A few things worth knowing
- Limits are per-server: one server's usage never affects another's.
-
Budgets are live, not a hard wall:
ctx.budget.runtimeMsRemainingandctx.budget.fetchCallsRemainingare exposed to your scripts so a data-dependent loop can size itself instead of guessing. - There are no separate caps on script runs or Discord API actions. Network calls simply run within your hourly script execution time.
- Founder is assigned by hand to the servers that supported Mallard early. There is nothing to buy, and no subscription to cancel.
Mallard is free forever. Nothing to buy, no subscription to cancel. If you'd still like to help with hosting, there's a tip jar on Ko-fi.