How to use AnimateDiff for Free

Time Needed : 00 hours 30 minutes

In this tutorial, we’ll guide you through the process of using AnimateDiff to create your own animated content, in less than 30 minutes. Let’s get started!

1. Step 1: Access AnimateDiff

   Go the the Replicate version of the project as discussed above. When there, you should see an interface as shown below:
   
   

2. Step 2: Inputting Main Workspace Parameters

   Input the required parameter within their appropriate text fields. The required parameters are as seen below:
   
   
   –motion_module: The motion_module, refers to a component or feature within the animatediff website that handles the animation and movement aspects of the generated images. The input type required here is a string (series of characters).
   
   –path: The path, refers to the trajectory or course followed by an animated element within the generated images. It involves the movement or transition of objects, text, or other visual elements over a series of frames or steps. The input type required here is a string (series of characters).
   
   –prompt: The prompt, is a user-provided input that serves as the basis for generating the content of the animated images. You may input text, ideas, or instructions, and the animatediff system utilizes this prompt to create visually engaging animations. The input type required here is a string (series of characters).
   
   –n_prompt: The n_prompt, refers to an extended or refined version of the prompt. It involves additional specifications or details provided by the user to guide the generation of the animated content. Additional specifications here involve what the system should not generate. The input type required here is a string (series of characters).
   
   –steps: The steps, represent the discrete stages or frames through which the generation process evolves. Each step contributes to the progression and refinement of the animated image. The input type required here is an integer (series of numbers).
   
   –guidance_scale: The guidance_scale, denotes a parameter or setting that influences the level of direction or control applied to the AnimateDiff generation process. It determines how closely the generated content adheres to the user’s input. The input type required here is a floating point (series of decimal or fractional numbers).
   
   –seed: The seed, is a starting point or initial configuration that influences the randomness or variability in the generated images. You might use seeds to achieve specific outcomes or styles in the animation. The input type required here is an integer (series of numbers).

3. Step 3: Inputting a Text Description for Your Video

   Ok, after glancing through your AnimateDiff interface and how it works, we can now proceed to input the required data for each field and generate our video. First, think of what you want to generate then type your idea into the prompt text field.
   
   
   For me, I use the following prompt: “masterpiece, best quality, 1girl, solo, cherry blossoms, hanami, pink flower, white flower, spring season, wisteria, petals, flower, plum blossoms, outdoors, falling petals, white hair, black eyes” as shown above.

4. Step 4: Tweaking Visual Effects

   After entering your prompt, for you to get the most out of your video, enter all the other extra parameters required.
   
   
   
   For me, I used the following parameters, as shown in the images above:
   
   –motion_module: mm_sd_v15_v2
   
   –path: realisticVisionV40_v20Novae.safetensors
   
   –prompt: masterpiece, best quality, 1girl, solo, cherry blossoms, hanami, pink flower, white flower, spring season, wisteria, petals, flower, plum blossoms, outdoors, falling petals, white hair, black eyes
   
   –n_prompt: badhandv4, easynegative, ng_deepnegative_v1_75t, verybadimagenegative_v1.3, bad-artist, bad_prompt_version2-neg, teeth, worst quality, low quality, nsfw, logo
   
   –steps: 25
   
   –guidance_scale: 7.5
   
   –seed: 255224557

5. Step 5: Previewing the Video

   Hit the Run button, at the bottom right of the left pane to assess the video’s alignment with your intent. The button is as shown in the image below:
   
   
   After clicking the “Run” button watch the terminal on the right pane, giving you extra information about your generation process, as shown below:
   
   
   Repeat the process, tweaking and re-arranging your prompt and parameters, until you get
   the desired output.
   
   NOTE: As of the time of writing this article, you have just 5 tries to generate your video, before the website blocks the generation option and asks for payment to continue.

6. Step 6: Exporting the Video

   Once satisfied, share or export the video as shown in the image below: