How to Use Suno: Full Workflow from Signup to Export

Updated 2026-09-13 · Suno-API team

New to Suno? Follow the full workflow step by step: sign up, pick a mode, write prompts, wait for generation, preview and export your track.

Most people who open Suno for the first time get stuck on step one: with all those input boxes on screen, which one do you actually fill in? The whole process is only five steps, and once you walk through it you're done learning.

Step 1: Create an account

Register on the site first. You only need a username and password - no email verification, no phone number. Once registration is complete, your balance and available models load automatically, and you don't need to configure an API key separately.

Step 2: Choose a mode

If you're not sure which to pick, start with simple mode. The system generates lyrics, melody, arrangement and vocals from your description, and produces two tracks at a time.

Step 3: Write your input

For simple mode, write the description in the order of style + theme + mood + instruments, for example: city pop style, a song about walking by the sea on a summer evening, light mood with a touch of nostalgia, lead instruments are electric piano and saxophone.

In custom mode you write the lyrics yourself. Mark sections with square brackets, such as [Verse] and [Chorus], so the model can tell verses from choruses more accurately.

Step 4: Submit and wait

After you submit, the task enters a queue. You can usually play it within 1 minute and download it within 5 minutes. You don't need to refresh the page during this time, and the task keeps running even if you close the page - you'll find it in your song list when you come back.

The same input produces two different renditions each time. The two may finish at different moments, so listen to whichever is ready first.

Step 5: Preview and export

• Preview online first and pick the one you're happy with.
• For everyday use, download MP3 - smaller files, faster uploads.
• For post-production or archiving, download WAV - higher audio quality but noticeably larger files.
• Neither format costs extra, and downloads are unlimited.

The four mistakes beginners make most

1. Writing only "a good song" as the prompt: that's too little information, so the model can only guess, and the result is usually not great.
2. Putting lyrics in the description box: the description box handles style, and lyrics have their own input field. Putting them in the wrong place affects the result.
3. Scrapping everything after one disappointing listen: listen to both tracks first - most of the time one of them is usable.
4. Downloading without checking the status: you can't get the file while the task is still processing. Wait until the status shows completed before you download.

Pricing

0.6 yuan per run, and each run always generates 2 tracks with unlimited downloads. There's no subscription and no credit expiry - you pay for what you use. If a generation fails, the cost is refunded after it's confirmed.