What's live
Ophraxx AI is now available for any Discord server to add and configure. The bot ships with Ophraxx Lite • OX-L1 — our fast, free-tier model built for everyday chat, quick questions, Discord moderation guidance, and short-form tasks. It runs through our own AI infrastructure, tuned for the low latency a chat environment demands.
Everything works out of the box once a server admin runs /setup. The interactive setup flow takes less than two minutes — admins pick an AI channel (or let the bot create one), toggle AI on or off, and choose whether @mention invocation is allowed outside the channel. Configuration is saved immediately and takes effect on the next message.
Safety and moderation from launch
We made a decision early on that safety could not be a post-launch add-on. Every message that reaches Ophraxx AI in beta passes through a five-layer pipeline: pattern-based pre-screening, an AI safeguard model covering nine threat categories, an automated fact-checker, a second safeguard pass on the output, and a final PII redaction step. None of these layers can be disabled by server admins or users.
The bot also ships with its own spam detection. Rapid message bursts trigger automatic soft-blocks and attempt to apply Discord's built-in timeout feature to the offending member. A detailed embed is sent to the server's mod log channel whenever a moderation action fires, so staff always know what happened and why.
Usage limits and what's coming next
Beta launches with enforced rate limits: 25 AI requests per user per day and 3,000 per guild per month. These exist to keep the platform stable while we monitor real-world load patterns. Server admins can check live usage at any time with /ophraxx usage. As usage patterns stabilize, we expect to revisit and raise these limits.
The beta period is also when we're gathering feedback through the thumbs-up and thumbs-down buttons on every AI response. This signal directly informs how we evaluate response quality across different server types and personality configurations. Ophraxx Lite • OX-B12 — the upcoming premium reasoning model — is already in development and will be the first major feature addition after launch.
