OpenAI Moderation is one model.
OpenModeration is the platform.
OpenAI's Moderation API gives you toxicity scores. OpenModeration gives you its own moderation AI, cross-checked by other models, pre-filtering, automated rules, human review by our team and DSA-compliant audit trails. It is hosted in the EU, and you can export your data and leave whenever you want.
The OpenAI Moderation API in 30 seconds
- Price: free. OpenAI states that the moderation endpoint is free to use.
- Model:
omni-moderation-latest, for text and images. - Categories: 13, from harassment and hate to self-harm, sexual content and violence.
- What it leaves to you: rules, human review, explaining decisions to users and the DSA audit trail.
Why teams switch from OpenAI Moderation
OpenAI's API is a good moderation engine. A moderation service needs everything that goes around it.
You're locked into one model's accuracy
OpenAI Moderation works very well on English, less so on other languages or niche categories, and it gives you no second opinion. OpenModeration runs its own moderation AI and has delicate cases cross-checked by other AI models. When a better model comes out, we plug it in.
An API call is not a moderation workflow
A toxicity score from OpenAI is only the start. You still have to decide which score triggers which action, review grey-area cases, log every decision for audits and manage your team's access. OpenModeration handles all of this, and our team reviews the doubtful cases.
Every comment goes to the API without pre-filtering
The OpenAI endpoint is free, but every comment still makes a round trip to it, even obvious spam or forbidden words. OpenModeration's pre-filter blocks these before they reach any AI model.
DSA compliance requires more than an API
The EU Digital Services Act requires documented moderation decisions, appeal mechanisms, and transparency reports. A raw API response is not enough. OpenModeration gives you audit logs, review trails and exportable reports.
Same text moderation, more control
With OpenAI's API, you get scores. With OpenModeration, you get scores plus pre-filtering, configurable thresholds, automated actions, and a complete audit trail, all through the same API call.
Our own moderation AI, set up on your rules and your use cases.
A second opinion: other AI models cross-check delicate cases before a decision is applied.
No frozen model: better or cheaper models are plugged in as they come out, with nothing to change on your side.
# OpenAI Moderation API (direct)
const openai = new OpenAI({ apiKey: 'sk-...' });
const mod = await openai.moderations.create({ input: text });
// Result: OpenAI-specific format, 13 categories, no thresholds
# OpenModeration
const result = await fetch('https://api.openmoderation.com/v1/moderation', {
method: 'POST',
headers: { 'Authorization': 'Bearer om_key_...' },
body: JSON.stringify({
input: text,
categories: ['hate', 'harassment', 'sexual', 'violence', 'self-harm', 'spam'],
threshold: 0.5
})
});
// Plus: pre-filter, double threshold, auto-actions, audit trail, real-time dashboardOpenAI Moderation API vs OpenModeration
A side-by-side look at the features you need to moderate in production.
| Feature | OpenAI Moderation API | OpenModeration |
|---|---|---|
| AI engine | One OpenAI model | Our own moderation AI, cross-checked by other AI models on delicate cases |
| API format | Proprietary OpenAI format | Standardized format. When we plug in a better AI model, nothing changes in your code. |
| Pre-filtering | Not available. Every request hits the API | Built-in word filter blocks obvious content before any AI call |
| Decision logic | Binary flagged/not flagged per category | Dual threshold: auto-approve, grey zone reviewed by our team, or auto-reject |
| Moderation Console | API only. No dashboard, no review UI. | Full dashboard: every decision in real time, team roles, analytics |
| Automated rules | Must build your own integration | Included. Define rules once and the Action Engine enforces them automatically |
| Human review | Must build your own review UI | Doubtful cases handled by our team; you can step in or override at any moment |
| Audit trail | Limited to API logs | Full audit trail: every AI and human decision logged, exportable for DSA |
| Vendor lock-in | Full lock-in to OpenAI ecosystem | None. Model-agnostic engine, and you can export your data at any time |
| Data sovereignty | Data processed on OpenAI servers | Hosted in the EU. On-premise deployment possible on request. |
| Pricing | Free endpoint (scores only) | Free up to 1,000 comments/mo. Managed cloud from €99/mo (10,000 comments included). |
| Max input length | 32,000 characters | Long texts supported + batch processing for larger volumes |
Using OpenAI Moderation API alone
- ✕ Every comment goes to the API, even obvious spam or forbidden words
- ✕ Binary flagged/not flagged, with no grey zone for nuanced comments
- ✕ Build your own dashboard for human moderators to review edge cases
- ✕ Build your own rule engine to translate scores into actions
- ✕ Build your own audit logging system for DSA compliance
- ✕ No second opinion on delicate cases
- ✕ Vendor lock-in to OpenAI ecosystem
OpenModeration
- ✓ Obvious spam and forbidden words are blocked before any AI call
- ✓ Two configurable thresholds decide what is approved, rejected or reviewed by our team
- ✓ Nothing to do on your side: you see every decision in real time and can step in whenever you want
- ✓ Rules set with three tools (forbidden words, category thresholds and your own rules in plain language), with no code
- ✓ A full audit trail, exportable reports and configurable data retention
- ✓ Delicate cases cross-checked by other AI models
- ✓ No frozen model: better models plugged in as they come out, and you can export your data
Built for every content moderation use case
You can replace OpenAI's API with OpenModeration in under an hour.
Social Platforms
Moderate user comments and posts with context-aware AI and community-specific policies.
AI Applications
Filter both user inputs and LLM outputs. Define your own categories as they evolve.
Gaming
Real-time chat moderation across multiple languages with sub-250ms latency and automated actions.
Marketplaces
Flag spam listings, hate speech in reviews, and PII leaks before they go live.
Moving from omni-moderation-latest to OpenModeration
Keep your integration simple and get the whole workflow around the score.
How the OpenAI Moderation API works: categories, pricing and limits →
Replace one call
One POST /v1/moderation call replaces openai.moderations.create, with a standardized response format.
Cross-checked decisions
On delicate cases, other AI models arbitrate the decision, and doubtful cases are handled by our team.
An AI that keeps improving
When a better or cheaper AI model comes out, we plug it in quickly. Nothing changes on your side.
OpenAI Moderation alternative: FAQ
Is the OpenAI Moderation API free?
Yes. OpenAI states that its moderation endpoint is free to use. It returns a flagged boolean and a score per category. Deciding what to hide, reviewing borderline cases and keeping an audit trail are left to you.
What does omni-moderation-latest detect?
omni-moderation-latest is OpenAI's current moderation model. It accepts text and images and scores 13 categories: harassment, harassment/threatening, hate, hate/threatening, illicit, illicit/violent, self-harm, self-harm/intent, self-harm/instructions, sexual, sexual/minors, violence and violence/graphic.
How much does OpenAI Moderation API cost vs OpenModeration?
OpenAI's moderation endpoint is free, but it only returns scores: no dashboard, no rules, no human review and no audit trail. OpenModeration has a free plan up to 1,000 comments analysed a month; the managed cloud starts at €99/month for 10,000 comments.
Which AI does OpenModeration use?
Its own moderation AI, set up on your rules and your use cases. On delicate cases, other AI models arbitrate the decision before it is applied. The engine is model-agnostic: when a better or cheaper model comes out, we plug it in, with no code to rewrite on your side.
How long does migration take?
Under an hour. One POST /v1/moderation call replaces openai.moderations.create. Scores map to a standardized format, and thresholds and rules apply immediately.
What do I gain beyond raw scores?
You get the rest of the workflow: a pre-filter, dual thresholds (auto-approve, review by our team, auto-reject), a dashboard where you see every decision in real time, team roles and a DSA-ready audit trail with exportable reports.
Which moderation API is best for enterprise compliance?
For enterprise compliance, you need more than raw scores: documented decisions, role-based access, audit trails and exportable reports. OpenModeration covers all of it. Every AI and human decision is logged, rules are enforced automatically, and the audit trail exports in one click for DSA transparency reports. EU hosting keeps data under your control, and on-premise deployment is possible on request.
Do I have to change my code when OpenModeration changes AI models?
No. Responses keep the same standardized format. When a better or cheaper AI model comes out, our team plugs it in quickly, and nothing changes on your side.
Compare other alternatives
Moderate with more than a single API.
Get our own cross-checked AI, pre-filtering, automated rules and human review by our team, hosted in the EU. You see every decision in real time.
No credit card required · DSA-compliant · GDPR-compliant
Questions? contact@openmoderation.com