Getting started with MaxVideo AI

1. Create your individual account and sign in to open the MaxVideoAI workspace.
2. Top up your wallet from Billing through secure Stripe Checkout. Pay by card; Apple Pay, Google Pay, PayPal, or Link may also appear when available for your device and region.
3. Choose a model and prepare your input. Add a prompt and, when the selected model supports it, a reference image. Review the displayed price before starting the generation.

Choose how you want to create

• Set up a generation yourself: use the Video, Image or Audio generator when you want to choose the model, inputs and settings directly.
• Get help with a prompt: open Milo inside that generator. Applying its suggestion fills your draft; it does not start a paid generation. Explore Milo.
• Develop a project through conversation: use Studio to work from a brief, organize references and prepare individual media. Assistance credits and media generation have separate costs; review each quote before approving it.
• Work from an external assistant: use a supported MCP connection with your MaxVideoAI account. Tools, attachments and approval steps depend on the host; check the connection guide.

Use the same account to find the media saved by these supported creation paths. Studio’s timeline and individual media downloads do not imply an export of an assembled film.

First generation checklist

1. Open the Video or Image workspace and select a model that supports your input type.
2. Confirm the duration, resolution, optional references, and the Price Before You Generate amount.
3. Start the generation, follow its status in Jobs, and find completed output in your library.

Video references can guide a new creation; they are distinct from source-video editing. Choose only the operations offered by the exact model, and check their source requirements before a quote. Luma Ray 3.2 Extend requires an eligible generation from your own account; FLUX 3 Draft Enhance also needs its qualified paired cache. A public video copy supplies neither private prerequisite. H3 Max Recast and Relight are specialized source tools; the new H3 Max Extend, Insert and Reframe remain pending.

Wallet and refunds

Confirming a paid generation charges your individual wallet. If the generation fails technically, the charge is automatically refunded in credits. The job remains visible in your history, including its status.

MCP availability

The public MCP guide explains how to connect a supported client and authorize your MaxVideoAI account. Available tools and generation approval steps depend on the client. Review its supported connection path before starting.

Let Milo help with the prompt

An idea is enough to start. In the Video, Image or Audio generator, open AI assistant and describe what you want. Milo uses your selected model, current settings and available reference context to help prepare a prompt. Read it, adjust it if needed, then choose Use prompt. You still decide when to generate, at the displayed media price.

Sign in to chat. Milo Lite is free within request and service limits. Standard and Max share 500 included assistance credits each month, with optional purchased credits. Video and audio excerpt analysis is separately quoted and confirmed. If credits run out, your draft stays: choose Lite or add credits.

See Milo in action.

Create audio on its own

Open Audio and choose the result you need:

• Voice over: write the exact spoken script, then choose a voice and the settings offered by that voice model. An audio reference is available with Seed Audio; it is not an input shared by every voice model.
• Instrumental: describe the genre, instruments and mood without vocals. Choose the available duration and tempo.
• Song: provide musical direction and lyrics, with verses and chorus on separate lines. The duration follows the generated song; an exact length is not offered.
• Sound effects: describe an event, its texture and how it unfolds, then choose a duration.
• Ambiences: describe a continuous environment and choose a duration. A seamless loop is not guaranteed.

Review the current price after filling in the inputs and settings, then start the generation. When it completes, listen in Audio, download the original file and reopen it from recent creations. Reuse settings is available when that result has a saved settings snapshot. Output format depends on the selected model and its offered options; this is not a universal MP3/WAV export selector.

Standalone audio is separate from sound generated natively by a video model. To add sound to an existing clip, Soundtrack a video opens a separate workflow: choose a source from your account or upload one, then review the available sound, music or voice options and the price. That workflow produces a video with its soundtrack; a separate audio file is an optional output when offered. It does not edit Studio’s whole timeline.

From a reference to a clip

1. Prepare an image you can use. For a recurring character, Character Builder can help prepare the reference.
2. Open Video, choose a model that accepts an image reference, add that image and describe the action. Check the offered format, duration and price; Milo can help refine the prompt before you generate.
3. Follow the job, open the completed result in Media, then download its original or choose a compatible next action.

Find and reuse your media

Media is your account’s library, not the public Examples gallery. Filter images, videos or audio, search for a saved item, or switch to creation history to check a job. Open an available output to preview it and download the original. Prompt, model and settings information depend on what was recorded for that result.

To create again, use Reuse settings where offered, or select the media as a reference in a compatible generator or tool. Check the receiving model’s inputs: reusing a prompt or media does not guarantee that every former setting or reference is restored. Review the new price before generating. Read the library overview.