Toxly

Moderation your product can actually use.

Toxly classifies user-generated text and returns a decision your app can act on: allow, warn, review, block, mask, or escalate.

Create a project Read the quickstart

Built for developers testing content safety in apps, communities, AI products, and Discord servers.

API keys Project policies Moderation logs Usage limits Discord bot
The product

A safety layer between user input and your app.

Toxly is not a replacement for judgement, support teams, or policy work. It gives your product a consistent first pass: score the text, apply your thresholds, store a safe log, and let your backend decide what happens next.

The dashboard is built around projects. Each project has its own API keys, policies, usage limits, logs, members, and integrations.

Text moderation API

Send text from comments, chats, profiles, prompts, support forms, or listings. Toxly returns structured JSON instead of vague labels.

Policy engine

Thresholds turn category scores into actions. You can tune strictness per project.

Dashboard

Review logs, create keys, manage members, edit policies, and test requests before shipping changes.

Discord moderation

Connect a server, choose actions per decision, and send moderation logs where your team works.

How it works

Four steps, no mystery box.

  1. 01

    Create a project

    Use one project per product, environment, community, or customer.

  2. 02

    Generate an API key

    Keys are shown once, stored hashed, and can be revoked from the dashboard.

  3. 03

    Send text

    Your backend calls Toxly and receives category scores, a risk score, matched rules, and a decision.

  4. 04

    Route the decision

    Allow safe content, warn users, queue review cases, block high-risk content, or escalate sensitive cases.

Developer surface

Small request. Stable response.

Use Toxly from your backend. The response is designed to be boring in the best way: predictable fields, clear categories, and decisions that map to product behavior.

curl -X POST https://api.toxly.net/v1/moderate/text \
  -H "Content-Type: application/json" \
  -H "X-Toxly-Key: txly_xxxxx" \
  -d '{
    "text": "message from your app",
    "metadata": {"source": "comment"}
  }'
Categories

Scores you can inspect.

Toxly exposes the category scores that shaped the decision, so your team can adjust policies instead of guessing.

toxicityhateharassmentviolence self_harmsexualminor_safetyspam scampiijailbreakprompt_injection
Logging

Useful logs without keeping the whole message.

Storedrequest ID, decision, risk score, category scores, matched rules, latency, text hash, and short preview.

Not stored in moderation logsthe complete original text.

Why it mattersteams can debug policy behavior without turning moderation logs into a content archive.

Plans

Simple limits for early teams.

Start with the free plan while integrating. Move up when request volume grows.

Free0 EUR1,000 requests / month
Starter9 EUR50,000 requests / month
Pro29 EUR250,000 requests / month
Scale79 EUR1,000,000 requests / month
BusinessCustomvolume by agreement
Current status

Built for first real testers.

Toxly is in an early production-ready beta. Core API moderation, dashboard projects, policies, logs, usage, project members, support tickets, and Discord setup are already in place. Image moderation and deeper model routing are planned next.

Create a project and send the first request.

Start in the dashboard, then follow the quickstart when you are ready to wire Toxly into your backend.

Create account Open docs