A Lens+ AI feature, announced September 21, 2026, that turns the text of a Snapchat chat message into a short, genre-styled AI song you can preview and send inside the conversation.
Last verified · 2026-09-22 · by Moe Ameen
Snapchat Chat to Song is an in-app AI feature that generates a short song from the text of a chat message. Snap announced it on September 21, 2026 as part of a wave of AI additions for Lens+, a paid tier within Snapchat+. The workflow lives entirely inside a conversation: open a chat, press and hold a message you sent or one a friend sent you, tap "Generate Song," choose a genre pill, and generate. When it is ready you can pick album art, preview the song, and tap "Send Sound" to share it in the thread, where it plays with an "AI Song" label.
It shipped alongside other Lens+ AI features announced the same day — AI Photoshoot (photo cleanup for blinks, expressions, and lighting), AI Remix for Snaps, Sticker Remix, and AI Fonts. Snap framed Chat to Song as a response to the short-form trend of turning conversations into songs. The company has not disclosed which model powers the music, whether it built the system in-house or licensed it from an AI-music provider, or what that model was trained on. The feature is rolling out in markets where Lens+ is available, access may vary by region, and it may not appear in every chat or account during testing; generation can also fail on a safety check, a connection issue, or a daily or rate limit.
Be clear on the scope: Chat to Song is a messaging feature, not a music-production or content tool. There is no full lyric, arrangement, or stem editing, and — more importantly for a creator — no documented way to export the song or post it to a public feed. It is designed to make one message land inside a Snapchat thread, and it stays there.
The catch with Chat to Song is that everything good about it — the song, the genre spin, the surprise — is sealed inside a Snapchat thread. There is no export and no path to a feed, so the moment you want an audience beyond the person you messaged, the feature has nothing for you. If an idea is fun enough to turn into a song, it is worth a real post; building that post is where [Kompozy](/) picks up. Kompozy does not generate the song for you — that is not its job — but it turns the same idea into content you can actually publish and own.
Concretely: take the line or moment you would have "song-ified" and bring it into Kompozy as one source. Kompozy generates the formats a public feed rewards — a captioned [Persona Short](/glossary/persona-shorts) fronted by a face-locked avatar, [Clipped Shorts](/glossary/clipped-short) cut from a longer video, a Marketing Short with a music bed, brand-exact [Carousel Posts](/glossary/hyperframes), Photo Posts, Quote Graphics, plus a Text Post, a Blog Article, and an Email Newsletter — all governed by one [Persona Brief](/glossary/persona-brief) so the voice and look stay consistent. Then [Autopilot](/glossary/autopilot) schedules and publishes the batch across the eight social platforms plus blog and email, each behind a per-post review gate. Snap gives Lens+ subscribers a private novelty; Kompozy is how you take the same creative impulse and make it work in public, on every platform, as content that is yours.
It is a Lens+ AI feature announced September 21, 2026 that turns a text message into a short AI song. In a chat you press and hold a message, tap Generate Song, choose a genre, preview, and send it into the conversation, where it plays with an "AI Song" label. It requires a paid Lens+ subscription.
Open a chat, press and hold the message you want, tap Generate Song, pick a genre pill and generate. When the song is ready you can choose album art, preview it, and tap Send Sound to share it in the thread. It works on messages you sent or received. Generation can fail on a safety check, a connection problem, or a rate limit.
Not directly. The song is created and shared inside a Snapchat conversation with an "AI Song" label, and there is no documented export or cross-posting path. To publish content around the text-to-song trend on public feeds, you make your own post — a content engine like Kompozy generates captioned video, carousels, and posts from one idea and publishes them across nine platforms.
Snap has not disclosed it. The company has not said whether the music model is its own or a licensed partner's, or what it was trained on — an open question given ongoing AI-music licensing disputes. Treat any specific claim about the model as unconfirmed until Snap details it.