Best Content Moderation APIs in 2026: Image, Video and Text Compared

Best Content Moderation APIs in 2026: Image, Video and Text Compared

01Introduction

If people can upload things to your app, someone will eventually upload something they shouldn't.

• A shopping app gets a nude photo in a product listing.
• A dating app gets abusive messages.
• A community app gets a violent video.
• A school platform gets bullying in the comments.

You can't check all of it by hand. A content moderation API does it for you. You send it an image, a video or a piece of text, and it tells you if the content is explicit, violent, hateful or toxic. Your app then blocks it, blurs it, or sends it to a person to review.

This guide compares the 7 best content moderation APIs in 2026: what they check, what they cost, and which one fits your app.

Quick picks

• Free: OpenAI Moderation (text and images)
• Biggest free tier: Azure AI Content Safety (5,000 images a month)
• Cheapest image moderation at scale: Amazon Rekognition ($1 per 1,000 images)
• Video moderation at scale: Amazon Rekognition or Hive
• Simple pricing for small teams: Sightengine
• Image, video and text moderation on one key, with an EU region: ApyHub

02What a content moderation API returns

You send content and get back a score for each risk. Here's a real response from ApyHub's Image Moderation API for a photo of a cat:

{
  "adultScore": 0.0007,
  "racyScore": 0.0014,
  "goreScore": 0.0006,
  "isAdultContent": false,
  "isRacyContent": false,
  "isGoryContent": false
}

A score near 0 means safe. A score near 1 means risky. You choose the cut-offs, for example:

• Above 0.8: block it automatically
• 0.5 to 0.8: send it to a person
• Below 0.5: let it through

03Content moderation APIs compared

Prices from each provider's pricing page, October 2026. ApyHub prices are worked out from its Pro and Pro+ plans.

04How to choose a content moderation API

Ask yourself five questions:

1. What do users upload? Images, video, text, or all three? Few APIs cover all three.
2. How much? At 1 million images a month, $1 per 1,000 is $1,000. $3 per 1,000 is $3,000.
3. What do you need to catch? Nudity is standard. Violence, hate, self-harm and drugs vary between providers.
4. Where is the content processed? User uploads are personal data under GDPR. If your users are in Europe, check where the API runs.
5. Do you need human review? Most teams block the obvious cases and send the unclear ones to a person. That needs scores, not just a yes or no.

05How video moderation works

Video moderation APIs don't watch the whole video. They take frames, for example one frame every second, and check each frame like an image.

That affects two things:

• Cost. More frames means more checks. A 10-minute video sampled once a second is 600 frames.
• Accuracy. Sample too rarely and you can miss a short scene. Sample too often and you pay for frames that look the same.

Amazon Rekognition charges per minute of video instead of per frame, which makes costs easy to predict. Sightengine lets you choose how many frames per second to check. Hive's price depends on the sampling rate you set. ApyHub's Video Moderation API uses Azure's or Google's own sampling.

06The 7 best content moderation APIs

1. OpenAI Moderation API

Checks text and images for 13 categories, including harassment, hate, self-harm, sexual content and violence.

• Good: completely free, and strong on text.
• Not so good: no video. Some categories, like hate and harassment, only work on text.

2. Google Cloud Vision SafeSearch

An image moderation API that rates images on five categories: adult, spoof, medical, violence and racy.

• Good: 1,000 free images a month, then $1.50 per 1,000, dropping to $0.60 above 5 million.
• Not so good: images only. Video and text need other Google services.

3. Amazon Rekognition

AWS's image and video moderation API, with a detailed list of moderation labels.

• Good: the cheapest image moderation at scale at $1 per 1,000 images, and video at $0.10 per minute. New AWS accounts get a 12-month free tier.
• Not so good: no text moderation, and it works best if you already use AWS and store your media in S3.

4. Azure AI Content Safety

Microsoft's text and image moderation API, with severity levels for hate, sexual content, violence and self-harm.

• Good: the biggest free tier: 5,000 images and 5,000 texts every month.
• Not so good: no video moderation, and paid prices only appear in the Azure calculator.

5. Hive

A moderation specialist with very detailed content classes.

• Good: $3 per 1,000 images and $0.50 per 1,000 texts, plus more than $50 in free credits.
• Not so good: self-serve plans are limited to 100 requests a day, and video needs a sales call.

6. Sightengine

A focused moderation API for images, video and text, popular with smaller teams.

• Good: simple pricing: 2,000 free operations a month, then $29 for 10,000.
• Not so good: the free plan caps you at 500 operations a day.

7. ApyHub moderation APIs

ApyHub covers images, video and text on one subscription:

• Image Moderation API: an NSFW detection API with adult, racy and gore scores, using ApyHub's own detector, Google or Azure.
• Video Moderation API: detects explicit content in videos through Azure or Google.
• Text Toxicity API: returns toxic or not, with a score. In our test, "You are an idiot and nobody wants you here." came back toxic with a score of 0.99.
• Good: one key for image, video and text moderation, plus spam detection, validation and file APIs. Dedicated EU region, and AI agents can call every API.
• Not so good: images cost more per check than AWS or Google at high volume, and video moderation needs your own Azure or Google account.

Try the moderation APIs free →

07Perspective API is shutting down

If you use Google Jigsaw's Perspective API for text toxicity, you need a replacement. Perspective is shutting down after December 31, 2026, and Google isn't offering migration help.

The closest Perspective API alternatives in this list:

• OpenAI Moderation: free, with harassment and hate categories
• Azure AI Content Safety: severity scores for hate and violence
• Hive text moderation: $0.50 per 1,000 texts
• ApyHub Text Toxicity API: a toxic score like Perspective's, on the same key as your other APIs

Start testing now, so you can compare results on your own data before the deadline.

08Content moderation in the EU

If your users are in Europe, where their uploads go matters.

• Azure and Google let you pick EU regions.
• AWS processes media in the region where your S3 bucket is.
• Sightengine is a French company.
• ApyHub is Dutch, and its EU region (api.eu.apyhub.com) keeps requests in the EU.

09Content moderation for AI agents

AI agents now read and post user content too: support tickets, uploads, replies. Through ApyHub MCP, Claude, ChatGPT, Cursor or Mistral Le Chat can call the image, video and text moderation APIs as tools. An agent can check an upload before it summarizes it, or check its own reply before it posts it.

Connect ApyHub MCP →

10Conclusion

There's no single best content moderation API. It depends on what you moderate and how much of it.

• OpenAI: free text and image checks
• Azure: the biggest free tier
• Amazon Rekognition: the cheapest images and video at scale
• Hive: detailed classes
• Sightengine: simple pricing
• ApyHub: image, video and text moderation on one key, an EU region, and moderation your AI agents can call

Start free on ApyHub, no card required →

11FAQ

What is a content moderation API?

A service that checks images, video or text for unsafe content like nudity, violence, hate or harassment. It returns scores or labels your app can act on.

Is there a free content moderation API?

Yes. OpenAI's Moderation API is free for text and images. Azure AI Content Safety includes 5,000 images and 5,000 texts a month, and Google Cloud Vision includes 1,000 images.

What is the best image moderation API?

Amazon Rekognition is the cheapest at scale, Azure has the biggest free tier, and OpenAI is free. Choose based on volume and where your data needs to stay.

Which content moderation APIs support video?

Amazon Rekognition, Hive, Sightengine and ApyHub's Video Moderation API. OpenAI and Azure AI Content Safety don't moderate video.

What is an NSFW detection API?

An image or video moderation API focused on adult content. Every image API in this list detects it.

What is the best Perspective API alternative?

OpenAI Moderation is the closest free option. Hive, Azure AI Content Safety and ApyHub's Text Toxicity API are paid alternatives with toxicity scores.

Can AI agents use content moderation APIs?

Yes. With ApyHub MCP, Claude, ChatGPT, Cursor and Le Chat can call moderation APIs directly as tools.

12About ApyHub

ApyHub is a curated API catalog and the trusted operational layer between your application and the services it depends on. One subscription, billed in atoms, covers the whole catalog of over 1,500 endpoints and capabilities, and it keeps growing. Every service is evaluated before it is listed and carries machine-readable certification covering GDPR, SOC 2 and ISO 27001. Every endpoint is MCP-ready by default. ApyHub runs dedicated EU and US regions and serves more than 65,000 developer workspaces every month. Start on the free tier with no card, or publish your own API on ApyHub.

13Sources

• OpenAI, Moderation guide
• Google Cloud, Cloud Vision pricing
• AWS, Amazon Rekognition pricing
• Microsoft, Azure AI Content Safety pricing
• Hive pricing · Sightengine pricing
• Jigsaw, Perspective API
• ApyHub, pricing