# audioworker full product reference > audioworker is a music studio you talk to. Bring a song, a photo or just an idea, say what you want in your own words, and the chat makes it: AI songs with lyrics in nine languages, songs from pictures, stems split from any track, AI vocals and choirs, mixing, effects, key and tempo changes, loudness, mastering and downloads of exactly what you hear. Audio Worker AI (audioworker) is a product and brand of J3X Solutions LLC. This file summarizes the public audioworker website. The linked pages remain authoritative; plans and limits can change, so use the features page and the policy pages when a current decision matters. ## Plans - Free ($0): 20 splits a day, one YouTube link a day, 2 GB, tracks kept 24 hours after your last activity. - Studio ($9.99 a month): 500 minutes of splitting and 100 YouTube links a month, 60 HD and AI minutes a month for AI songs, vocals and masters, 50 GB, kept until you delete them. - A YouTube link also uses one split. There is no yearly plan. AI songs, picture to song, AI vocals, choirs, instrument parts, singing in your own voice, HD splits, mastering and HD masters are on Studio. ## AI song maker audioworker writes a whole song from a description typed into the chat. The description can be as loose or as exact as you like: a mood, a story, a genre, a tempo, a voice. Lyrics come from you, pasted with structure tags such as [Verse] and [Chorus], or the AI writes them from the idea. Lyrics can be in English, Korean, Japanese, Chinese, Spanish, French, German, Portuguese or Italian; by default they follow the language of your last message. The voice can be male, female, a duet or any voice, or the song can be instrumental. A song runs from 10 seconds to five minutes, about two minutes if you do not say. A reference song, from a YouTube link, an upload or a track already in your library, shapes the tone and the mix of the new song without copying it. When the song comes back you listen to the versions, keep one or all, and the kept song becomes an ordinary track: split it into stems, change the mix, add effects, master it and download it in the same conversation. AI songs are on Studio and use its HD and AI minutes (60 a month). What you can say: - A bossa nova about my cat who steals socks, sung as a duet. (The AI writes the lyrics and two voices sing them.) - Use my lyrics. Slow piano ballad, female voice, in Korean. (Your words, sung in Korean.) - Make it feel like this track, but it's about my hometown. (The reference shapes the sound. It is never a cover.) - Three minutes of rainy day lo-fi, no vocals. (An instrumental at the length you asked for.) What comes back: - Your lyrics or AI lyrics: Paste yours, or the AI writes them. - Nine languages: Lyrics follow the language you write. - Any voice, or none: Male, female, a duet, or instrumental. - A song to keep shaping: Keep one, then split, mix, master. Plans: Free: AI songs need Studio. Studio ($9.99 a month): 60 HD and AI minutes a month. Questions: - How long can a song be? From 10 seconds to five minutes. Two, if you don't say. - Can I use my own lyrics? Yes, if they are yours or licensed. Or the AI writes them. - Who owns the song? You keep your rights. Nothing you make trains AI. ## Picture to song Picture to song turns photos into music. Drop one to ten pictures into the chat in one message, add a line about the mood, the language, the voice or the length if you want, and audioworker writes a song whose story and lyrics come from what the pictures show. Studio makes up to 20 picture songs a day from its HD and AI minutes. The song arrives like any AI song: listen, keep a version, then split it into stems, mix it, add effects, master it and download it. Your pictures are private to your account and are never used to train models. What you can say: - Three beach photos: "A summer song to sing in the car." (A song built on what the pictures show.) - My grandmother's garden: "Gentle folk, for her birthday." (Your line steers the song the photo starts.) - My dog at the window: "Punk song about the mailman." (The story comes from the picture, the attitude from you.) - A city at night: "In Spanish, female voice, two minutes." (Language, voice and length as you asked.) What comes back: - One to ten pictures: Dropped into one message. - Your direction, or none: Mood, language, voice, or nothing at all. - Up to 20 songs a day: On Studio, from HD and AI minutes. - Then keep going: Split, mix and master the keeper. Plans: Free: Picture songs need Studio. Studio ($9.99 a month): Up to 20 songs a day. Questions: - Which plan do I need? Studio. Sign in, then drop the photos into the chat. - Are my photos used to train AI? No. Pictures, chats and audio never train models. - Can I change the song afterwards? Ask for another, or change its mix, key or effects. ## AI vocals, choirs and parts On a track that has been split into stems, Studio adds new parts by request. AI vocals sing your lyrics or lyrics the AI writes, in a female, male, duet or any voice. An AI chorus puts a female, male or mixed choir on up to four sections of the song, singing lyrics you write, the song's own hook or wordless ooh and aah. Instrument parts can be generated for one section or the whole song. A vocal stem can also be switched to a stock AI voice of the other gender, keeping the words and the melody. My voice lets a person sing in their own voice: record one take of 10 to 30 seconds and tick three consent boxes (it is your own voice, you are 18 or older, you agree to the processing). One voice per account; recording it is free and singing with it is on Studio. Imitating a real person's voice is not allowed. Each new part arrives as its own stem in the player, to keep, mix or drop. What you can say: - Write a chorus about leaving town and have a female voice sing it. (A sung take over your stems.) - A mixed choir on the last two choruses, just ooh and aah. (A wordless choir on those sections.) - Switch this vocal to a male voice. (The same words and melody, a different voice.) - Sing it in my voice. (Your own recorded voice, after your consent.) What comes back: - AI vocals: Female, male, duet or any voice. - AI chorus: A choir on up to four sections. - Instrument parts: For one section or the whole song. - My voice: One take of 10 to 30 seconds. Plans: Free: Record your voice. Singing needs Studio. Studio ($9.99 a month): 60 HD and AI minutes a month. Questions: - Whose voice can I use? Only your own, with consent, if you are 18 or older. - Can the choir sing words? Your lyrics, the song's hook, or wordless ooh and aah. - Which languages can it sing? English, Korean, Japanese, Chinese, Spanish, French, German, Portuguese and Italian. ## Stem splitter audioworker splits a song into four stems that stay in time with each other: vocals, drums, bass and other. Bring an audio file (MP3, WAV, M4A, AAC, FLAC, OGG, Opus, WMA or AIFF), a video file (MP4, M4V, WebM, MKV, MOV or AVI) up to 400 MB and 90 minutes, or a YouTube link. You confirm you hold the rights before anything is processed. A link already in your library is reused without fetching it again. The stems open in a player under the chat. Each stem has its own volume up to 150 percent, mute, solo and a left or right speaker, changed by asking or by hand while it plays. Download any single stem, or the mix exactly as you hear it, as MP3 or WAV. Free splits 20 songs a day (a YouTube link counts as one of them, one link a day). Studio splits 500 minutes a month with 100 YouTube links, adds HD splits, and keeps tracks until you delete them. What you can say: - Split this and turn the bass up. (Four stems, then the bass up in the mix.) - Just the drums from 1:20 to 1:50, on repeat. (Drums soloed, that bar looped.) - Split this link, mute the vocals, give me an MP3. (The instrumental, as a file.) - Send me the bass stem as a WAV. (One stem, one file.) What comes back: - Four stems: Vocals, drums, bass and other, in time. - Files or links: Audio, video, or a YouTube link. - Mix while it plays: Volume, mute, solo, left or right. - HD splits: Higher quality separation on Studio. Plans: Free: 20 splits a day. Studio ($9.99 a month): 500 minutes a month and HD splits. Questions: - Is it really free? Yes: 20 splits a day, kept for 24 hours. - Which files can I use? Most audio and video files, up to 90 minutes. - Do I need the rights to the song? Yes. You confirm them before anything is processed. ## Vocal remover To remove the vocals from a song, bring it to the chat as a file or a YouTube link and ask. audioworker splits it into vocals, drums, bass and other, mutes the vocals and plays the instrumental; ask for "only the voice" and it solos the vocals for an a cappella instead. The vocals can also be turned down rather than off, or muted only for a section. The key can move up or down six semitones without changing the tempo, so a karaoke track fits your voice, and the speed can go from half speed to 1.2 times without changing the pitch. The Download button renders the mix exactly as you hear it, as MP3, WAV or an MP3 at -14 LUFS, for the whole song or one section. Free covers 20 songs a day. What you can say: - Take the vocals out so I can sing over it. (The instrumental, ready to sing over.) - Only the voice, please. Nothing else. (The a cappella, soloed.) - Two semitones down so it fits my voice. (Lower key, same tempo.) - Download this karaoke version as a WAV. (The mix exactly as you hear it.) What comes back: - Instrumental: Every part but the voice. - A cappella: The voice on its own. - Your key: Six semitones either way, same tempo. - Download as heard: MP3 or WAV, whole song or a section. Plans: Free: 20 songs a day, with downloads. Studio ($9.99 a month): 500 minutes a month and HD splits. Questions: - How do I make a karaoke track? Say "remove the vocals", listen, then ask for the download. - Can I keep a little of the singer? Yes. Turn the vocals down, or mute just one part. - Does a new key change the speed? No. The key moves, the tempo stays. ## Mixing and mastering audioworker mixes by conversation. Each stem has a volume up to 150 percent, mute, solo and a left or right speaker; a 10-band EQ boosts or cuts each band by up to 12 dB; and reverb, delay, drive, filter, chorus, wah and lo-fi can be placed on any section of any stem. Sections can be muted, faded, cut, moved and undone, stems duplicated or renamed, your own recording added as a stem, and a mix saved by name to come back to. For the finish, a loudness-matched file at -14 LUFS (streaming), -16 (podcasts) or -9 (clubs) works on every plan; mastering and HD masters are on Studio. The Download button renders exactly what you hear, as MP3, WAV or an MP3 at -14 LUFS, for the whole song or one section. What you can say: - Vocals up a little, drums down, bass only on the left. (The mix changes while it plays.) - Reverb on the vocals in the bridge only. (An effect on just that section.) - Make it ready for streaming. (A file at -14 LUFS.) - Master it in HD and send me a WAV. (A Studio master, lossless.) What comes back: - Each stem's level: Up to 150 percent, mute, solo, pan. - 10-band EQ: Up to 12 dB on every band. - Effects on a section: Reverb, delay, drive, chorus, wah, lo-fi. - Loudness for the place: Streaming, podcast or club levels. Plans: Free: Mixing, EQ, effects and loudness. Studio ($9.99 a month): Mastering and HD masters. Questions: - Does the download sound like what I hear? Yes: levels, edits, EQ and effects, as MP3 or WAV. - Loudness or mastering? Loudness works on Free. Mastering is on Studio. - Can I undo a change? Yes, and a mix can be saved by name. ## Key and tempo changer For practice, audioworker moves the key of a song up or down six semitones without changing its tempo, and the speed from half speed to 1.2 times without changing its pitch. Any section can be looped by its start and end times. Split the song first and every stem can be soloed or muted, so you can practise against the band without the part you play. Ask for the key, the tempo or the chords and the chat reads them from the song and draws the chords along the waveform. A click track arrives as its own stem, muted until you turn it up, and a part can be written out as MIDI. What you can say: - Down a whole step, I can't reach the high notes. (Two semitones lower, the same tempo.) - Slow it to 70 percent and loop the solo. (Slower, the same pitch, the solo on repeat.) - What key is this in, and what are the chords? (Key, tempo and a chord line under the waveform.) - Add a click so I can play along. (A click track, muted until you turn it up.) What comes back: - Key up or down: Six semitones, the tempo unchanged. - Speed: Half speed to 1.2 times, same pitch. - Loop a part: Set the start and end, repeat. - Chords, key and tempo: Read from the song, plus a click. Plans: Free: Key, speed, loops and 20 splits. Studio ($9.99 a month): Every HD and AI tool. Questions: - Does slowing down change the pitch? No. Speed and key are separate. - Can I practise one instrument? Split the song, mute your part, loop the hard bit. - How do I find the key of a song? Ask. The chat reads the key, tempo and chords. ## Rights, privacy and support You keep your rights in what you upload and what you make, AI-generated audio included, and audioworker claims no ownership. You confirm you hold the rights to a song before it is processed. Your audio, pictures, chats and outputs are never used to train AI models. Support: support@audioworker.ai. Policies: https://audioworker.ai/terms, https://audioworker.ai/privacy, https://audioworker.ai/dmca.