Quick Start

This quick start introduces the Flova homepage and project interface, shows how to start a video project from different entry points, and explains how to keep editing, generating, working in parallel, and exporting your final content.

Start with the page tour to learn where common entry points are and what each area does after you enter a project.

Page Tour

Global

• After logging in, enter a simple Prompt in the homepage input box to start a new project.
• View your credits in the top-right corner, and click to browse and choose a membership plan.
• Open Projects to view and reopen all historical projects, then continue creating.
• Open Skill to browse selected Skills from Flova and creators. A Skill is the Agent's operating guide. You can also write or revise your own Skill there.
• Open Flova TV to browse selected videos created by Flova users and view their creation process.

Inside a project

After opening any project, Flova follows a "direct from the right, view and edit on the left" workflow. Content completed by the Agent appears in the corresponding panel on the left, and each panel also supports further manual editing.

• Give tasks in natural language from the chat box on the right.
• At the bottom of each AI message, choose back to this moment or Brunch in new project.
• Use the storyboard to view key elements, shots, global music or voiceover descriptions, and generated media automatically attached under each shot.
• Use the Media Files Panel to browse all project media, including media already linked to the storyboard and media not yet linked.
• Use the timeline to view the result assembled by the Agent and preview the rhythm of the final video.
• Click any media item to show it in the Preview area. Below it, you can use comment generation, Edit, and regeneration.
• Any panel on the left can be manually edited and can work in parallel with Agent tasks on the right.

Start a Project

Start with one Prompt

Enter a short Prompt in the homepage input box to describe the video you want to make. In addition to the content topic, you can include basic requirements such as duration, aspect ratio, genre, and style. If you have images, videos, or audio that help express the idea, upload them as reference media and briefly explain how to use them. If you already have a song or BGM, upload it too and explain whether it should be used as the final soundtrack, as the rhythm and mood basis for a music video, or simply as inspiration for the visual style.

Examples:

Turn the story of Cinderella into an AI short film of about 3 minutes, horizontal 16:9, English dialogue, 3D animation style, with a light and upbeat rhythm.

Create a healing video of about 2 minutes: after closing the store, a night-shift convenience store clerk receives a note left by his late grandmother and decides to wait for someone outside a hospital the next day. Horizontal 16:9, Chinese dialogue, gentle realistic style.

Create a 30-second wireless earbud ad: fast cuts across subway, morning run, and office scenes, highlighting lightness and noise cancellation. Vertical 9:16, no dialogue, fast-paced BGM, polished commercial ad feel.

Use the song I uploaded as the final BGM and create a music video of about 3 minutes: psychedelic electronic, neon rainy-night city, visuals intensify with the chorus. Horizontal 16:9, no dialogue.

Start from a vague idea or story inspiration

You do not need a clear story at the beginning. If you have a BGM track, an inspiring reference image, or even a vague idea, you can ask the Agent to expand it into story concepts first. Upload an image or music file, or describe your idea in chat, and clearly ask the Agent not to generate any actual content yet. Brainstorm the plot and design the storyboard structure first, then move into generation after the direction is confirmed.

Examples:

I uploaded this BGM and want to make a music video. Do not generate anything yet. First, brainstorm several story premises based on the lyrics and style of the music.

I want to create an ad for a sports earbud brand. Do not generate anything yet. First, brainstorm several 30-second vertical ad concepts and storyboard directions.

Start from a script or shot list

If you already have a clear script or detailed shot list, paste or upload the script, synopsis, storyboard table, or shot list, and let the Agent organize it and move the project forward. If you do not want the storyboard to change your script or shot descriptions, state clearly in the Prompt that your text should not be rewritten.

Example:

Refer to the text file I uploaded. It contains a detailed script and shot list. Do not modify the content in any way. After organizing it, do not generate anything yet. Wait for me to confirm the storyboard.

Start from existing media assets

If you already have character design images, scene images, background music, or existing media such as product logos and product images for an ad project, upload them together. If you do not want the Agent to regenerate these characters or scenes, state that directly in the Prompt.

Example:

Refer to the script I uploaded, along with the character design images, scene images, character voice files, and background music. Organize the storyboard according to the script. Use the characters, scenes, and character voice files I provided directly without regenerating them. Do not rewrite the script or modify the references. After generating the storyboard, wait for my confirmation.

Start from a media task

Not every creation needs to become a full video. Many projects begin with media generation, such as testing model results with an existing video Prompt in Seedance 2.0, generating a piece of music first, or creating a character image. Creation itself can be exploratory.

In the project, tell the Agent directly, for example: "generate a piece of music like this", "generate images of this character", or "use this Prompt with Seedance 2.0 to generate a video". The media will first enter the Media Files Panel, because no storyboard is planned at this stage. Whenever you are ready to start producing the final video, ask the Agent to create the storyboard.

Examples:

Use Seedance 2.0 to generate a 15-second video, and do not create a storyboard yet. Prompt: "0-5s: wide shot of a rainy night street, neon reflections on wet pavement, camera slowly pushes forward. 5-10s: medium tracking shot, a girl holding a red umbrella walks past lit shop windows. 10-15s: close-up side profile, water droplets and sign reflections sweep across her face, cinematic feel."

Based on the person in the image I gave you, I want to generate several versions in different styles. Do not create a storyboard yet.

Based on the script I uploaded, find the key characters and scenes and generate images for them. Do not create a storyboard yet.

Start by recreating a video

If you saw a short film or ad online and want to learn from its shot structure, rhythm, or camera movement to make your own version, upload that video and explain in the Prompt what you want to reference. Usually, recreation means learning from structure and rhythm, while the visual content is changed to your own topic or media.

Example:

I uploaded this reference video. I want to follow its shot structure and editing rhythm to create my own brand ad. My product reference is in the image I uploaded. Vertical 9:16, about 30 seconds.

Import existing media and only complete later steps

If you already have live-action footage, a rough cut, or clips produced in other tools, you can only supplement AI shots and assembly. Import images, videos, and audio into the project, or explain the purpose of each media item in chat. Tell the Agent which shots need to be newly generated.

Start from a Skill

On the homepage or Skill page, you can find classic or distinctive Skills created by users. Choose a Skill to start a project, add your topic or media in the input box, and let the Agent execute according to that Skill. This is useful when you want to first experience what kind of video the Skill can produce before deciding whether to use it in your own work.

Start from Flova TV

Flova TV contains many selected finished videos, along with the creator's process and the Skill they used. You can start a project directly with the same Skill attached to a finished video, or watch videos for inspiration and return to the homepage to create your own project with a Prompt.

Continue from a historical project

• Open any historical project from Projects on the homepage and continue from its current state.
• At the bottom of each AI reply, use Brunch in new project to explore different directions and versions within the same project.

Edit Your Project

Use Add to Chat for precise references

Usually, you can describe what you want to change directly in the chat box. If the target is not clear enough, add the specific object to chat first, then give your instruction. Specific content in the storyboard, Media Files Panel, timeline, and documents can all be referenced. The Agent will use the full information of the referenced object to understand exactly what you want to modify.

• Reference storyboard content: key element, shot, and audio cards all support references. Move your mouse to the upper-left handle of a card and click Add to Chat to bring the full information of that element, shot, or audio into the chat.
• Reference media assets: choose Add to Chat from the handle in the bar below an asset. For a single media item, right-click and choose Add to Chat, or click Add to Chat in the upper-right corner of the Preview area. Both assets and individual media items can also be dragged directly into the chat box as references.
• Reference text content: in documents or storyboard descriptions, select the exact text you want to modify and click the floating Add to Chat action to bring that text into the chat.

Edit project content

Besides asking the Agent to handle changes, you can directly operate on the storyboard, documents, and media assets in the interface. For media organization, small text mistakes, or issues where you already know the correct fix, manual editing is often more direct and helps bring the current version back on track faster.

• Edit the storyboard: double-click a text area in the storyboard to enter edit mode and directly change key element, shot, or audio descriptions. You can also add elements, shots, and global audio. Each type of storyboard card has a handle in the upper-left corner. Drag it to reorder cards, or delete entries you no longer need.
• Adjust media asset ownership: drag the handle in the bar below an asset in the storyboard to change which storyboard category it is attached to, or remove it from the storyboard.
• Organize media: assets support basic actions such as delete and download. Individual media items also support like, delete, and download. You can drag a media item into another asset as a new version, or create a separate asset from that media item.
• Add media: areas that store assets, such as the Media Files Panel and storyboard, usually have a + entry. Click it to upload media or enter Edit.
• Edit: the timeline supports basic editing operations. You can manually adjust clip order, trim clip length, and make basic rhythm changes to the final video.
• Edit documents: documents usually have an Edit button in the upper-right corner. Click it to modify the document content directly.

Edit generation settings

In the generation area below the Preview area, click Edit to view the Prompt, parameters, reference images, and reference media used by the current media asset. You can modify the Prompt directly, adjust parameters, or replace reference media. Reference media can be dragged into this area from panels or anywhere else to add new references or replace existing ones.

Comment generation

Comment generation is also located below the Preview area's generation section. It is useful when you want to describe how the current media should change: simply explain the desired change in natural language, and the Agent will continue generating new results based on the current media.

Parallel generation

Manual edits to any module, as well as Edit and comment generation, can run in parallel with Agent tasks in the chat box on the right. For example, you can let the Agent continue generating the next batch of new content while using comment generation or Edit on the left to modify and regenerate earlier media. This avoids waiting for one path to finish before doing anything else and improves creation efficiency.

When parallel generation causes manual edits on the left and Agent edits on the right to affect the same content, the system opens a conflict panel and asks you to choose which version to keep. After confirmation, the project continues from the version you selected.

Export a project

When the project is complete, use the Export button in the top-right corner to choose different export formats: export the final video, export PR-format files, or export all files in the project.