Can Content Moderation check images?

Yes. Send an image_url block with a public http(s) URL, alone or alongside text blocks. Content Moderation applies the self-harm, sexual and violence families to images, except sexual/minors; the remaining categories are text-only. Base64 and data: URIs are not accepted — host the image and pass its URL.

How do I read the Content Moderation result?

Each entry in moderation.results has flagged (true when any category is triggered), categories (per-category booleans), category_scores (0–1 confidence) and category_applied_input_types. Treat flagged as the model's default policy and the scores as the knob for your own — a stricter product can hold anything above a lower threshold for review.

Can I moderate several texts in one Content Moderation request?

Yes. Pass an array of strings and Content Moderation returns one result per string, in order, from a single task with a single bill. Pass an array of text and image_url blocks instead when the items belong together and should be judged as one submission. Each request takes up to 500 items and 30,000 words of text in total; split larger jobs across requests.

How do I get the result back?

Content Moderation is async on reApi. POST to /api/v1/moderations and you get a task id; poll /api/v1/tasks/:id until status is completed, then read the moderation object from the output. Content Moderation checks usually finish in a few seconds; polling adds no extra charge.

Does Content Moderation require a monthly subscription?

No. Content Moderation pricing is pay-as-you-go: $1 buys 1,000 credits and credits never expire. Cost scales with the words and images you check rather than with a recurring fee.

How do I get an Content Moderation API key?

Sign up and create a key in the dashboard, no card required. Pass it as a Bearer token in the Authorization header on the /api/v1/moderations endpoint, and use the same key for every other reApi model.

Content Moderation — Screen Text & Images Against 13 Harm Categories

Q: Which categories does Content Moderation detect?

Harassment, harassment/threatening, hate, hate/threatening, illicit, illicit/violent, self-harm, self-harm/instructions, self-harm/intent, sexual, sexual/minors, violence and violence/graphic. Each comes back as a boolean plus a 0–1 score, and Content Moderation also lists which input types — text, image — each category was evaluated against.

Content Moderation (model id omni-moderation-latest) checks text, images, or both against 13 harm categories and returns a flagged verdict with per-category scores. Send content, poll the task, act on the result — over the same async /api/v1 contract you already use.

Highlights: 13 harm categories · text, image, or both
Best for: Screening user content before it goes live
Input: Text, image URLs, or a mix
Output: Flagged verdict + per-category scores

Input

Plain text · or a JSON array of strings / content blocks · up to 500 items and 30,000 words per request

Image URLs

Drag images here or click to upload

0/16 files

Each image counts as one unit

Estimated cost~$0.0011 credits

Result

Try one of these prompts

Pricing

Credit-based — 1 credit = $0.001 USD. Pay only for completed generations.

First test cost

$0.0011 credits

One sample at the cheapest tier (1,000 words).

Testing budget guide

Add credits

$10

≈ 10000 tests

$50

≈ 50000 tests

$100

≈ 100000 tests

Text: $0.001; Price · 1,000 words
Image: $0.001; Price · image

What is Content Moderation on reApi?

Content Moderation is the omni-moderation-latest model: it takes text, image URLs, or both in one request and returns a flagged verdict with 13 per-category scores and the input types each category was applied to. One async call handles a single string, a batch of strings, or a mixed text-and-image submission, and every request is billed per moderation unit — 1,000 words of text or one image.

What you can build with this model

Real-world workflows and production use cases you can build and ship with this model.

Content Moderation screening a stream of community messages through a shield gate

Moderate a live community feed with Content Moderation

Comments, chat messages, forum posts and reviews arrive faster than any team can read them. Content Moderation screens each one against harassment, hate, violence, sexual content, self-harm and illicit categories and returns a flagged verdict with a score per category, so your pipeline auto-approves the obvious, holds the borderline for a human, and blocks the clear violations. Batch a whole page of comments into one call — Content Moderation returns one result per string.

Check some content

Content Moderation evaluating a photo and its caption inside one lens

Screen images and captions together before they publish

A photo and its caption can be harmless apart and abusive together. Content Moderation accepts a mixed input — text blocks plus image URLs in one request — and scores the combination as a single item, reporting which categories were evaluated against the image and which against the text. Put Content Moderation in front of your upload endpoint so user-generated media never reaches the CDN unchecked.

Content Moderation sorting a grid of content cards into cleared and held lanes

Backfill and audit content at pipeline scale

Re-screen an archive, audit a partner's catalog, or QA the output of your own generation models. Content Moderation bills per moderation unit — 1,000 words of text or one image — and rounds the bill up once per request, so large batches cost less per item than single calls. Same task id, same poll endpoint, same credit ledger as every other reApi model, so Content Moderation drops into an existing worker without new infrastructure.

Why reAPI for Content Moderation

13 categories, one verdict

Content Moderation reports harassment, harassment/threatening, hate, hate/threatening, illicit, illicit/violent, self-harm, self-harm/instructions, self-harm/intent, sexual, sexual/minors, violence and violence/graphic — each as a boolean plus a 0–1 score — and a single flagged decision on top. Tune your own thresholds on the scores when the default verdict is too coarse.

Text, images, or both in one request

Send a string, an array of strings, or an array of text and image_url blocks. Content Moderation tells you which input types each category was applied to, so a mixed submission is scored as one item rather than two disconnected checks. Content Moderation takes image URLs only — no base64 uploads to manage.

Priced per unit, rounded once

1 credit equals $0.001 USD. Content Moderation bills one unit per 1,000 words of text (each text item counts as at least 50 words) and one unit per image, then converts the total to credits once per request. Failed requests are refunded automatically; polling is free.

Content Moderation vs Azure AI Content Safety

Both screen text and images for harmful content. They differ in how many categories they report, how a result is shaped, what a billable unit is, and how much setup sits between you and the first call.

Capability

Content Moderation on reApi

Azure AI Content Safety

Integrate Content Moderation in three steps

step 01
Create an API key
Sign up at reApi and grab a key. Signup credits cover plenty of Content Moderation checks — enough to validate the verdicts against content you already know the answer for.
Open
step 02
Submit content to check
POST to /api/v1/moderations with model: omni-moderation-latest and your input — a string, an array of strings, or text and image_url blocks. Content Moderation returns a task id immediately.
Open
step 03
Poll for the verdict
GET /api/v1/tasks/:id until status is completed. The response carries moderation.results — flagged, categories, category_scores and category_applied_input_types for each input item.
Open

docs/api/content-moderation

API reference

Drop-in code and the full parameter table.

View full reference

curl -X POST https://reapi.ai/api/v1/moderations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "omni-moderation-latest",
    "input": [
      { "type": "text", "text": "Is this comment safe to publish?" },
      { "type": "image_url", "image_url": { "url": "https://example.com/photo.jpg" } }
    ]
  }'

Parameter validation failed

Rejected synchronously with the failing field named. Check enum values, ranges, and URL fields — nothing is charged.

Authentication or balance issue

401 means a missing or invalid Bearer key; 402 means the reserve exceeds your balance. Manage keys and credits in the console.

Content or material rejected

Reference material and outputs pass automated moderation. Rejected tasks fail with a clear error and a full refund.

Task failed or timed out

A task that reaches a failed state is never charged — the reserve refunds automatically. Keep the task id and retry when ready.

Frequently asked questions

Common questions about this model.

Content Moderation is the omni-moderation-latest model exposed on reApi: it checks text, images, or a mix of both against 13 harm categories and returns a flagged verdict plus a score per category. Call it on the /api/v1/moderations endpoint with model id omni-moderation-latest; the verdict comes back as moderation.results over an async task, billed per moderation unit. See the current rate in the pricing table on this page.

Credit-based: 1 credit equals $0.001 USD. Content Moderation bills one unit per 1,000 words of text — summed across every text item, each counted as at least 50 words — plus one unit per image, with a minimum of one unit per request. The unit total is converted to credits once and rounded up, so batches cost less per item. See the current per-unit rate in the pricing table on this page. Failed requests are refunded automatically.

Paid on reApi: completed checks are charged at the per-unit rate in this page's pricing table, with no subscription. The model's creator, OpenAI, offers its own moderation endpoint free to OpenAI API accounts; on reApi you get Content Moderation with the same key, ledger and task contract as every other model here. Eligible new accounts may receive promotional test credits; technical or provider failures are refunded automatically.