# AI music generator: make songs and background music from text

Describe a style or paste your lyrics, pick a length, and download a royalty-free stereo MP3 made with Google Lyria 3.5.

## How it works

1. Describe the genre, mood, instruments, and scene in up to 2,000 characters—or press Use an example or a style tag like Lo-fi, Cinematic, or Acoustic.
2. Pick a length (30 s with Lyria 3 Clip, or 1, 2, or 3 min with Lyria 3.5), keep Instrumental on or switch to With vocals to add your own [Verse] and [Chorus] lyrics, and set a BPM if you want one.
3. Generate the track, preview it in your browser, and download the 44.1 kHz stereo MP3 (plus timed .txt lyrics)—or press Open in editor to mix it under your voice.

## What you can compose with Audo's AI music generator

Generate instrumental beds that sit cleanly under dialogue, 30-second intros and jingles, or full multi-minute songs from your own lyrics:

### Podcast intro & theme music: Royalty-free podcast intro music, outro themes, and background beds

Skip generic stock loops that ten other shows already use. Pick 30 s to compose original podcast intro music or a gentle outro cue, then send the MP3 to Mix audio to automatically duck the music 12 dB whenever a host or guest speaks and normalize the episode to podcast loudness (see our podcast loudness LUFS guide).

### YouTube & video background music: Instrumental music generator for YouTube videos, Shorts, and streams

Leave Sound on Instrumental so no vocals clash with your narration. Match the tempo to your edit—from slow ambient pads to upbeat lo-fi or electronic beats—and drop the track onto an MP4 or MOV in Replace audio in video, or trim and layer it alongside captions in Audo's online video and audio editor.

### Lyrics to song & AI song maker: Turn lyrics into a song with AI vocals, verses, and choruses

Use Audo as an AI song generator and AI song maker for lyrics to song: switch Sound to With vocals and paste up to 3,000 characters of your own words with [Intro], [Verse], [Chorus], [Bridge], and [Outro] tags plus backing vocals in parentheses. Turn the finished track into a shareable waveform video in MP3 to MP4.

### AI jingle generator & music API: Make 30-second ad jingles in the browser or from AI agents

As an AI jingle generator and jingle maker, Lyria 3 Clip composes a catchy 30-second brand hook or sung slogan in ~10 seconds, ready for monetized social ads and product launches. Developers and AI assistants can automate text to music from ChatGPT, Claude, Claude Code, and Cursor via Audo's MCP server and REST API—see Audio and video tools for AI agents.

## Royalty-free text to music with Google Lyria 3.5—built into your editing workflow

Audo's online AI music generator and royalty-free music generator composes original 44.1 kHz stereo MP3 tracks from a written description up to 2,000 characters (and optional custom lyrics up to 3,000 characters) using Google's Lyria 3.5 for full 1-to-3-minute songs (~30 seconds) and Lyria 3 Clip for 30-second intros and jingles (~10 seconds). You can generate 2 full tracks a day free with no sign-up—cleared for commercial and monetized use—and run extra tracks pay-as-you-go at $0.10 per clip or $0.20 per full song from $5, with no monthly subscription.

Specific musical directions produce much stronger arrangements than one-word moods. Name the genre or era, the lead and backing instruments (such as warm acoustic guitar, Fender Rhodes piano, upright bass, or swelling strings), the atmosphere, an optional root key like G major or D minor, and a tempo from 40 to 220 BPM. On Lyria 3.5, you can also guide transitions with timestamps—such as a quiet piano intro from 0:00 to 0:12 before drums enter—and specify vocal range (like female alto or male baritone) when singing is turned on.

Unlike standalone song generators that stop at a download button, every track on Audo opens straight into Audo's post-production tools: check its detected key and tempo in Key and BPM finder, isolate drums or instruments in Stem splitter, pair it with an AI voiceover from Text to speech, or trim it cleanly in Cut audio. Every track is freshly composed, royalty-free for YouTube, podcasts, and client work, and never claims exclusive Content ID ownership.

How Audo's AI music generator compares on free access, pricing, and commercial rights (checked November 2026)

| Tool & model | Free tier | Paid pricing | Commercial & monetized rights | Source |
|---|---|---|---|---|
| Audo AI (Google Lyria 3.5 & 3 Clip) | 2 full tracks a day, no sign-up (44.1 kHz stereo MP3) | Pay once from $5 ($0.10 per 30 s clip, $0.20 per 1–3 min song); no monthly subscription | Royalty-free commercial and monetized use on both free and paid runs | https://audo.ai/ai-music-generator?utm_source=markdown&utm_medium=agent |
| Suno (v6 & v6-mini) | Daily free allowance on v6-mini (account required) | Pro from $10/month ($8/month billed yearly); monthly allowance resets | Free plan is personal, non-commercial use only; commercial rights require a paid plan | https://help.suno.com/en/articles/9601601 |
| ElevenLabs Music (v2 / v2.5) | Monthly free allowance (~3–5 tracks, account required) | Starter $6/month, Creator $22/month; API $0.15/minute | Free plan is personal use only; paid plans include commercial rights (streaming from Creator) | https://elevenlabs.io/music |
| Google Gemini API (Direct Lyria 3.5) | No free tier on Lyria 3.5 or Lyria 3 Clip | Developer API billing ($0.08 per song, $0.04 per 30 s clip) | Google claims no ownership of output; includes SynthID audio watermark | https://ai.google.dev/gemini-api/docs/pricing |

Tool: Generate music (`generate_music`). Price: 10 to 20 credits per job — $0.20 per job, pay-as-you-go with no monthly subscription — per job’s measured quantity, charged only when a job succeeds.

## Free API

```bash
curl -F 'text=Calm lo-fi beat with soft piano, for a study video' -F duration_seconds=60 https://audo.ai/api/try/generate-music -D response.headers -o response.body
```

Read response.headers and response.body before naming media. Handle 202, 200 JSON and errors with the complete response client at https://audo.ai/docs/api?utm_source=markdown&utm_medium=agent#responses.

Free for 2 full tracks a day per person (30-second Lyria 3 Clip or 1-to-3-minute Lyria 3.5 song), with no key or account. Plain UTF-8 descriptions up to 2,000 characters (2 MB) and optional lyrics up to 3,000 characters. Free results last one hour. The response is the result itself; a job that runs longer than about 30 seconds returns 202 with a link to check on it. With an API key (`-H "Authorization: Bearer $AUDO_API_KEY"`), omitting free runs the full job with credits if the balance covers it, otherwise the free allowance if quota remains. Set free=true to prohibit spending; for paid runs use free=false with an approved positive max_credits. Details: https://audo.ai/docs/api.md

## Options

Form fields with the same names as the MCP tool's inputs:

- `text`: What the music should be, in words, up to 2,000 characters: genre, mood, instruments, and where it's used, such as "calm lo-fi beat with soft piano and vinyl crackle, for a study video". Lyrics go in lyrics. Pass text or text_file_id, not both.
- `model`: The music model and its price a job: lyria_3_5 (Lyria 3.5, 20 credits: full tracks with structure and an ending), lyria_3_clip (Lyria 3 Clip, 10 credits: quick 30-second clips for intros and loops). Default lyria_3_5.
- `duration_seconds`: lyria_3_5 only: about how long the track should be, 30 to 180 seconds, such as 60. A target the model aims for, not a cut. Default: the model decides, usually about 2 minutes. lyria_3_clip always makes 30 seconds.
- `vocals`: true for a song with a singer, false for instrumental music only. Default false, or true when lyrics are given.
- `lyrics`: Words to sing, up to 3,000 characters, with optional section tags such as [Verse] and [Chorus], and backing vocals in parentheses. Leave out with vocals true and the model writes them.
- `bpm`: The tempo in beats per minute, 40 to 220, such as 90. Default: the model picks one that suits the description.

## Languages

Lyria 3.5 and Lyria 3 Clip sing in the language the description and lyrics are written in (including English, Spanish, French, German, Japanese, Korean, Portuguese, and Hindi). Instrumental tracks have no words.

## From an assistant (MCP)

Add https://audo.ai/mcp and sign in (setup for each assistant: https://audo.ai/setup.md). The tool is `generate_music`:

> Composes new music from a description as an MP3, instrumental by default or sung with vocals and lyrics. lyria_3_5 (default) makes full tracks aimed at duration_seconds (30 to 180); lyria_3_clip makes a quicker 30-second clip. Artists' names and copyrighted lyrics are refused. Not for speech (text_to_speech) or mixing existing music (mix_tracks). It costs 20 credits a job (lyria_3_5) or 10 (lyria_3_clip), charged only if it succeeds; free twice a day. Returns the audio and any sung lyrics with download links, or job_id for get_job.

Example: "Make a 1-minute calm lo-fi track for my podcast intro."

## Questions

**Is there a free AI music generator with no sign-up?** Yes. Audo generates 2 full tracks a day free—either a 30-second clip or a 1-to-3-minute song—with no sign-up, no spoken watermark, and a 44.1 kHz stereo MP3 download. Unlike free tiers limited to personal listening, Audo's free and paid tracks are royalty-free for commercial use. When you need more, tracks are pay-as-you-go ($0.10 for a 30-second Lyria 3 Clip or $0.20 for a full Lyria 3.5 song, charged only when generation succeeds), with no monthly subscription—you pay once from $5 and your balance never expires.

**Is AI-generated music royalty-free and legal to monetize on YouTube, podcasts, and ads?** Yes. Music generated on Audo with Google's Lyria 3.5 and Lyria 3 Clip is royalty-free for commercial and monetized YouTube videos, podcasts, Twitch streams, client videos, and paid ads, with no attribution required. What no prompt-only AI music tool can grant is exclusive copyright ownership or YouTube Content ID registration: under U.S. Copyright Office guidance, purely AI-generated music cannot be copyrighted on its own, and the model may compose a similar melody for another prompt. Every track includes Google's imperceptible SynthID audio watermark identifying it as AI-generated.

**How do I turn my own lyrics into a song, or make purely instrumental music?** Switch Sound to With vocals on Audo and paste up to 3,000 characters in Lyrics using structure tags like [Intro], [Verse], [Chorus], [Bridge], and [Outro], plus parentheses such as (moving on) for backing vocals—or leave Lyrics empty and Lyria 3.5 writes and sings original words in the language of your prompt. For background music under speech, keep Sound on Instrumental so no vocals compete with your voiceover. Artist names, band styles, and copyrighted lyrics are declined automatically at no charge.

**What is the difference between Google Lyria 3.5 and Lyria 3 Clip, and how long can a track be?** Selecting 30 s on Audo runs Google's Lyria 3 Clip ($0.10 after free runs), which composes a fixed 30-second piece in ~10 seconds for podcast intros, ad jingles, and short loops. Selecting 1 min, 2 min, or 3 min runs Google's Lyria 3.5 ($0.20 after free runs), which arranges full multi-section songs up to about 3 minutes in ~30 seconds and finishes on a natural musical ending rather than cutting off mid-bar. To build a longer background bed, generate two matching tracks and join them in Merge audio.

**How do I make podcast intro music or duck background music under a voiceover?** Pick 30 s or 1 min on Audo, keep Sound on Instrumental, set a steady tempo like 90 or 110 BPM, and describe the instruments and tone—for example, "warm acoustic guitar and light brushed drums, welcoming podcast intro with a clean ending." Once the track is generated, press Open in editor or send the MP3 to Mix audio, which automatically ducks the music behind your spoken voiceover and brings it back up between sentences.

**How does Audo compare with Suno as a pay-as-you-go alternative?** Suno focuses on standalone consumer songwriting on a monthly subscription ($10/month Pro for commercial rights, while Suno's free tier is personal, non-commercial only) and offers no public developer API. Audo is built for creators, podcasters, and developers who want Google's Lyria 3.5 and Lyria 3 Clip with 2 free commercial-ready tracks a day, pay-as-you-go pricing from $5 with no monthly subscription, built-in voiceover ducking and video tools, and direct API and MCP access.

**Can ChatGPT, Claude, or coding agents generate music with Audo?** ChatGPT and Claude can write lyrics and chord ideas on their own, but they cannot render downloadable MP3 music tracks natively. Once you Add Audo to your assistant via Audo's MCP server (https://audo.ai/mcp), ChatGPT, Claude, Claude Code, Codex, and Cursor can call generate_music from a plain-English prompt, return the finished 44.1 kHz stereo MP3, and mix it under your voiceover or onto an MP4 video in the next step—or call the music API directly with curl.

## More

- Every tool: https://audo.ai/llms.txt
- OpenAPI: https://audo.ai/api/openapi.json
- This page for people: https://audo.ai/ai-music-generator?utm_source=markdown&utm_medium=agent