CleanMod

Perspective API shuts down DecemberΒ 31.

Switch in five minutes: change one URL.

CleanMod is a drop-in Perspective API replacement. It uses the same comments:analyze request and response format, so your moderation code keeps working after Google turns Perspective off.

Before and after

Before: Perspective API (shuts down Dec 31, 2026)
curl -X POST \
  "https://commentanalyzer.googleapis.com/v1alpha1/comments:analyze?key=$GOOGLE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"comment": {"text": "You are an idiot"},
       "requestedAttributes": {"TOXICITY": {}}}'
After: CleanMod
curl -X POST \
  "https://cleanmod.dev/v1alpha1/comments:analyze" \
  -H "Authorization: Bearer $CLEANMOD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"comment": {"text": "You are an idiot"},
       "requestedAttributes": {"TOXICITY": {}}}'

The body and the response shape stay the same. You can also keep sending the key as a ?key= query parameter, the way Perspective clients do.

Same request and response

comments:analyze with comment.text, requestedAttributes, languages, scoreThreshold, doNotStore, spanAnnotations and clientToken. Responses return attributeScores with summaryScore and spanScores.

Same errors

Google's error envelope with Perspective's documented messages, so your existing error handling keeps working.

Works with Google's client libraries

CleanMod serves a Perspective discovery document, so the googleapis Node client and the Python discovery client work once you point them at cleanmod.dev.

Attribute coverage

Perspective's six production attributes, plus SEXUALLY_EXPLICIT. The _EXPERIMENTAL variants return the same scores. English only for now; need another language? Tell us which.

Perspective attributeLanguages on CleanMod
TOXICITYEnglish
SEVERE_TOXICITYEnglish
IDENTITY_ATTACKEnglish
INSULTEnglish
PROFANITYEnglish
THREATEnglish
SEXUALLY_EXPLICITEnglish
Scores won't match Perspective's numerically. CleanMod uses open models trained on Jigsaw's public comment datasets, not Google's production models. Re-check your thresholds on a sample of your own comments before you switch. See all differences.
πŸ”Œ

CleanMod for WordPress

The official CleanMod WordPress plugin adds toxicity detection on top of WordPress' own comment moderation. It complements Akismet: Akismet fights spam, CleanMod fights abuse and toxicity.

  • βœ“Checks new comments automatically
  • βœ“Flags or blocks toxic and abusive submissions
  • βœ“Respects your existing moderation rules
  • βœ“Configure it under Settings β†’ CleanMod in wp-admin
View WordPress plugin β†’

Ways to use CleanMod

Perspective-compatible API

Available now

WordPress plugin

Available now

JavaScript SDK

Available now

REST API

Available now

Not migrating from Perspective?

CleanMod's own API is a few lines of code too

/api/v1/moderate returns a single allow, flag or block decision with category scores. Read the docs.

cURL
curl -X POST https://cleanmod.dev/api/v1/moderate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "You are an idiot"}'

Start free, scale when you need to

CleanMod has a free plan, so you can integrate and check scores on your own comments before you cut over. Move to a paid tier when you need more volume.

View pricing

Get your replacement key before December 31

Create a free account, generate an API key, and point your Perspective client at cleanmod.dev.

Get a free API key