The model makes the clip. Video API makes it ready to ship.
The catalog
* FAMILY · minimax/minimax

MiniMax 2.5 Music and Text-to-Image

This MiniMax family pairs the 2.5 music model, which turns lyrics and style prompts into produced tracks, with a realistic text-to-image generator.

2 variants$0.15 per runminimax provider
* MINIMAX VARIANTS
2 of 2
minimax minimax music 2.5 Music Generation preview
Music Generation
minimax music 2.5
minimax · minimax
Minimax 2.5 is an advanced music generation model that transforms lyrics and style prompts into fully produced audio tracks with cohesive composition, vocals, and instrumentation.
minimax Minimax | Text to Image Text to Image preview
Text to Image
Minimax | Text to Image
minimax · minimax
Minimax Text to Image is an advanced generative model that transforms written prompts into highly realistic images. It focuses on clarity, vivid details, and natural lighting to create visually appealing results.
* ABOUT MINIMAX

Working with MiniMax

Two unrelated capabilities share this family. MiniMax Music 2.5 turns lyrics and style prompts into fully produced audio with cohesive composition, vocals and instrumentation. MiniMax Text to Image generates realistic images with an emphasis on clarity, vivid detail and natural lighting.

When to choose MiniMax

Come here for MiniMax's image generator, or for the 2.5 music model specifically. The dedicated MiniMax Music family carries the newer 3.0 and is the better default for audio.

Choosing between the variants

Music 2.5 produces a complete track from lyrics and style direction. Text to Image is the still-image generator and shares nothing with it beyond the provider.

* FREQUENTLY ASKED

About MiniMax

Should I use Music 2.5 or MiniMax Music 3.0?
3.0 in the dedicated music family is newer and accepts lyrics, prompts or instrumental direction. Use 2.5 when its particular arrangement style is what you want.
What does MiniMax Text to Image focus on?
Clarity, vivid detail and natural lighting — it leans photographic rather than illustrative.
Do the two variants share any settings?
No. They are separate models filed under one family slug, so treat them as unrelated endpoints.