[{"data":1,"prerenderedAt":115},["ShallowReactive",2],{"content:\u002Fguides\u002Fvoice-message-moderation":3,"ArticleBody_NrFeuk5ZszHlD5t5u93QGqhrwABihfHpyBNtDg4":110},{"kind":4,"path":5,"locale":6,"title":7,"htmlTitle":8,"description":9,"navTitle":10,"published":11,"updated":11,"alternates":12,"breadcrumbs":13,"indexable":21,"twin":22,"collection":23,"slug":24,"file":25,"content":26,"hub":19,"titleEm":28,"eyebrow":29,"lead":30,"cta":31,"faq":34,"faqJsonLd":53,"sources":60,"related":86,"topic":106,"jsonLdType":107,"blocks":108,"shortDefinition":109,"feed":21},"article","\u002Fguides\u002Fvoice-message-moderation","en","How to moderate voice messages, to the second","Voice Message Moderation, to the Second · Detectivision","Check voice messages for profanity, insults and hate speech, with the start and end second of each finding: how it works, what to do next, which APIs do it.","Voice message moderation","2026-11-05T21:44:00+03:00",[],[14,17,20],{"name":15,"path":16},"Home","\u002F",{"name":18,"path":19},"Guides","\u002Fguides",{"name":10,"path":5},true,"\u002Fguides\u002Fvoice-message-moderation.md","guides","voice-message-moderation","content\u002Fen\u002Fguides\u002Fvoice-message-moderation.md",{"collection":27,"path":5},"guides_en","to the second","Guide","Send each voice message to a speech moderation API, which turns the speech into text with the time of every word and checks it for your categories. Each finding comes back with its start and end in seconds, so a moderator plays only that part and your app can mute or hold it. Here is how it works, what to do with a finding, and what each API publishes.",{"label":32,"to":33},"Audio moderation API","\u002Faudio-moderation",[35,38,41,44,47,50],{"q":36,"a":37},"Can an API detect abuse in voice messages?","Yes. A speech moderation API turns the recording into text with the time of every word and checks it by category. Detectivision's audio moderation checks speech for profanity, insults, hate speech, sexual content, self-harm and bad habits, and gives each finding with its start and end in seconds.",{"q":39,"a":40},"How do I find where in a recording the abuse is?","Use an API that returns the time of each finding. Detectivision gives start and end in seconds from the beginning of the recording, so a moderator plays only that part and your app can mute or cut it.",{"q":42,"a":43},"Which languages can voice messages be in?","Detectivision's audio moderation understands English, Spanish, Portuguese, French, German, Italian, Turkish, Arabic, Russian, Hindi, Chinese and Japanese, the same 12 languages as its text moderation.",{"q":45,"a":46},"How much does it cost to check a voice message?","On Detectivision a recording costs 1 audio credit per started minute of sound when the answer comes by webhook, and twice that when you wait for it. A 40-second voice message is 1 credit by webhook. One check takes up to 5 minutes of sound.",{"q":48,"a":49},"Should a flagged voice message be deleted automatically?","Only for the categories and confidence your rules are sure about. Hold the rest before delivery or send them to a moderator, who plays the flagged seconds and decides. Tell the sender why a message was held.",{"q":51,"a":52},"Can the speech of a live stream be checked the same way?","Yes. Detectivision checks the speech of a live stream as it comes in, next to its frames; see live stream moderation.",[54,55,56,57,58,59],{"q":36,"a":37},{"q":39,"a":40},{"q":42,"a":43},{"q":45,"a":46},{"q":48,"a":49},{"q":51,"a":52},[61,65,68,71,74,78,82],{"title":32,"url":62,"publisher":63,"accessed":64},"https:\u002F\u002Fdetectivision.ai\u002Faudio-moderation","Detectivision","2026-11-05",{"title":66,"url":67,"publisher":63,"accessed":64},"Credits and billing (docs): audio, a unit per started minute of sound","https:\u002F\u002Fdetectivision.ai\u002Fdocs\u002Fcredits",{"title":69,"url":70,"publisher":63,"accessed":64},"Introduction (docs): what each moderation type takes and returns","https:\u002F\u002Fdetectivision.ai\u002Fdocs",{"title":72,"url":73,"publisher":63,"accessed":64},"Webhooks (docs)","https:\u002F\u002Fdetectivision.ai\u002Fdocs\u002Fwebhooks",{"title":75,"url":76,"publisher":77,"accessed":64},"Audio Moderation & Speech to Text","https:\u002F\u002Fdocs.thehive.ai\u002Fdocs\u002Faudiovideotext","Hive",{"title":79,"url":80,"publisher":81,"accessed":64},"Detecting toxic speech (Amazon Transcribe Developer Guide)","https:\u002F\u002Fdocs.aws.amazon.com\u002Ftranscribe\u002Flatest\u002Fdg\u002Ftoxicity.html","Amazon Web Services",{"title":83,"url":84,"publisher":85,"accessed":64},"Moderation (OpenAI API guide)","https:\u002F\u002Fdevelopers.openai.com\u002Fapi\u002Fdocs\u002Fguides\u002Fmoderation","OpenAI",[87,90,94,98,102],{"path":33,"title":88,"description":89},"Audio moderation","Speech in podcasts, voice messages and recordings is transcribed and checked for profanity, insults and hate speech, with the exact seconds of each.",{"path":91,"title":92,"description":93},"\u002Flive-stream-moderation","Live stream moderation","Moderate live streams and cameras as they air: RTSP, RTMP, HLS, SRT. Frames and speech are checked for your flags, paid for the frames checked.",{"path":95,"title":96,"description":97},"\u002Fuse-cases\u002Fdating-apps","Dating apps","Check profile photos, bios and chat messages on your dating app for nudity, harassment, hate and scams, in 12 languages, with one API.",{"path":99,"title":100,"description":101},"\u002Fuse-cases\u002Fgaming","Gaming","Check in-game chat, player names and uploaded images for insults, hate, sexism and sexual content, in 12 languages, with one API.",{"path":103,"title":104,"description":105},"\u002Fguides\u002Farabic-turkish-hate-speech-moderation","How to moderate hate speech and insults in Arabic and Turkish","Why English-first moderation misses Arabic and Turkish abuse, which APIs list both languages for which categories, and how to test one on your own messages.","by-content","TechArticle",false,null,["Island",111],{"key":112,"result":113},"ArticleBody_NrFeuk5ZszHlD5t5u93QGqhrwABihfHpyBNtDg4",{"head":114},{},1793904542816]