Suno Generate
Full song generation using Suno AI with extensive creative controls.
Overview
Suno Generate creates complete songs from text prompts. It supports multiple Suno model versions, custom lyrics with metatag formatting, genre/style tags, and fine-grained controls for vocal gender, style weight, weirdness, and audio weight. The prompt field supports Suno metatag autocomplete for structured lyrics.
Configuration
| Field | Type | Default | Description |
|---|---|---|---|
| Prompt | string | "" |
Main text prompt describing the song (in custom mode, the lyrics). Supports Suno metatag autocomplete. Max length is per-version: 5000 for V4.5 / V4.5PLUS / V4.5ALL / V5 / V5.5, 3000 for V4; 500 in non-custom (inspiration) mode. Over-long input is truncated to the model’s limit — the editor warns first so you can switch model or shorten. |
| Model | enum | "V5" |
Suno model version: V5, V4_5ALL, V4_5PLUS, V4_5, V4. |
| Title | string (max 80) | "" |
Title for the generated song (Suno caps titles at 80 characters). |
| Lyrics | string | "" |
Song lyrics with metatag support ([Verse], [Chorus], [Bridge], etc.). Same per-version max as Prompt (5000 for V4.5+/V5, 3000 for V4). |
| Style | string (max 1000) | "" |
Genre and style tags (e.g., “pop rock, upbeat, energetic”). Max 1000 for V4.5+/V5, 200 for V4. |
| Negative Style | string (max 500) | "" |
Styles to avoid in generation. |
| Vocal Gender | enum | auto | "male", "female", or unset for automatic selection. |
| Style Weight | number | 0.5 |
Influence of style tags on output (0.0 to 1.0). |
| Weirdness | number | 0.0 |
Controls experimental/unconventional output (0.0 to 1.0). |
| Audio Weight | number | 0.5 |
Balance between prompt and audio characteristics (0.0 to 1.0). |
| Custom Mode | boolean | false |
Enables advanced parameter control. |
| Instrumental | boolean | false |
When true, generates instrumental-only (no vocals). |
Inputs & Outputs
Inputs
| Handle | Accepts | Routes to |
|---|---|---|
| Prompt | text producers + audio/visual pickers | prompt |
| Audio style | audio pickers (genre/mood/instrumentation), voice describers | folded into style/prompt |
| Voice | voice persona (suno-voice / voice-design / voice-character) | personaId |
| Style (Advanced) | text producers | style |
| Lyrics (Advanced) | text producers | lyrics |
| Title (Advanced) | text producers | title |
| Negative style (Advanced) | text producers | negativeStyle |
The four Advanced handles appear when you expand Advanced ▾ on the node (or when the field already has content/a wire). Wiring any of Style / Lyrics / Title puts Suno into custom mode automatically.
Outputs
audio— generated audio URL
Bottom-strip controls
The node’s bottom strip exposes the most-used non-text controls without opening the config panel:
- Model — Suno model version.
- Instrumental —
With vocals/Instrumental — no vocals. - Vocal —
Auto/Male/Female(hidden when Instrumental is on). - Mix ▾ — three advanced sliders (0.00–1.00):
- Style Weight — how literally Suno follows your Style tags.
- Weirdness — how far Suno strays from genre norms.
- Audio Weight — how strongly a wired voice-persona / reference steers the song.
The same sliders (with these descriptions) are in the config panel.
Editing fields from the node
Each field — Prompt, Style, Lyrics, Title, Negative style — is filled one of three ways, with no manual source-picker:
- Type directly (with Suno tag suggestions), or
- Connect a node to the field’s input handle — its output is injected into that field automatically. A wired field shows read-only; unwire it to type by hand. (Wired input pips stay visible even when Advanced is collapsed, so you never lose track of a live connection.) Or
- Reference a
{variable}in the text — it resolves from the upstream node with that label at run time.
Expand Advanced ▾ and use Edit ▾ to edit any field — including Prompt — in a focused modal (the same editors as the config panel). The node’s Final view shows the fully-assembled result — prompt + style + lyrics + title + negative + any connected style pickers — so you always see exactly what will be sent.
Generate with AI
The Prompt, Style, Lyrics, and Negative style fields each have a ✨ AI button (in the config panel and the field-edit modal) that runs the prompt wizard:
- Style / Negative style → a comma-separated list of Suno style tags (genres, moods, instruments, production traits — for Negative style, traits to exclude).
- Lyrics → complete song lyrics with
[Verse]/[Chorus]/[Bridge]section tags — actual sung words, not tags or a prompt.
Each button is hidden when its field is wired from another node. (Title has no AI button — it’s a short label.)
Best Practices
- Use Suno metatags in lyrics (
[Verse],[Chorus],[Bridge],[Outro]) to structure the song. - Keep Style Weight around 0.5 for balanced results; push toward 1.0 only when you need strict genre adherence.
- Set Weirdness to 0.0 for predictable output; increase gradually for more experimental results.
- Use Negative Style to explicitly exclude unwanted genres (e.g., “metal, screaming”) rather than relying on the prompt alone.
- V5 produces higher quality output than V4 models.
Common Use Cases
- Generating a complete song from a text description and lyrics.
- Creating instrumental background music for video content.
- Producing genre-specific tracks (pop, rock, jazz, electronic) using style tags.
- Rapid prototyping of song ideas before professional production.
- Generating multiple variations of a concept by adjusting style weight and weirdness.
Tips
- Combine structured lyrics (with metatags) and a descriptive prompt for the most coherent results.
- The Instrumental toggle is useful for creating backing tracks that can later receive vocals via Suno Add Vocals.
- When using Custom Mode, all advanced parameters (style weight, weirdness, audio weight) become active.
- Connect a Suno Lyrics node upstream to auto-generate lyrics before song creation.