Character Creator Workflow | Runway Academy
Skills:
LLM Engineering70%
Key Takeaways
The video demonstrates the Character Creator workflow in Runway, a tool that utilizes AI to generate character images and videos. The workflow nodes are explained, and examples are shown on how to create new character poses and expressions.
Full Transcript
Welcome to Runway Academy. Today we're exploring a featured workflow called Character Creator. [music] Workflows is an amazing tool to use for character creation as it allows you to generate anything that you need all at once. Let's explore this character creator workflow, which you can access yourself in the featured workflows section of workflows. Our first node is the prompt for the character description, which then gets fed into an LLM node to be enhanced with some custom system instructions. The LLM node is helping to describe the character in more detail, as well as adding some rules, such as making sure the character is always pictured in a full body, neutral, forward- facing pose towards the camera with their whole face and front body visible. and the background always being described verbatim as a flat chroma green background color with no shadows. From there, our first character image gets created. That character image is going to serve as our global foundational image reference so that the character stays consistent in every single generation. The first set of images that get created are neutral expression, full body, and close-ups. The prompts for each of these is quite simple. From there, the character gets created in three distinct poses. T-pose, right profile, and left profile. These poses get generated with the help of a reference chart, which is plugged into image slot two, right underneath our global character reference image. Using charts or pose images in general, is a great way to get your character looking exactly how you want. Let's quickly demonstrate getting our character into a new pose. Drag your main character reference image onto an empty part of the canvas to create a new image reference node. Then right click on the canvas and add a text node, another image upload node, and an image generation node. Hook up the text node to the prompt input, your main character reference into the image one input, and the empty image node into the image 2 input. Now upload any pose reference image. Then in the prompt box, say something like, "Show me the character from image one in the same pose as the reference from image two." Make sure the body proportions from image one stay the same. Now, your character will be in a new custom pose. Going back to the main workflow, now we have some additional images being created that showcase various character expressions in both wide shots and close-ups. Surprised, angry, happy, bored, confused, and sad. All of the close-up images are using the same text prompt node as our original close-up image that got created with a simple text prompt of zoom in and frame this exact character in a close-up. Under all of those image nodes, we have some video key frame nodes that are using the wide full body shot as frame one and the close-up as the last frame, which creates a dynamic zoomin expressive video of your character. The prompt driving these nodes is quite simplistic and generic. The video naturally and seamlessly transitions from the first frame to the last frame without cuts, transitions, or inconsistencies. [music] You can, of course, be more detailed about this prompt if you'd like. Last but not least, we have some additional videos being created. The first is a 360° character rotation plate. This node is very valuable if you need to extract any angle of your character. [music] Simply pause the video at the frame that you need, right click on the node, and choose save video frame as to download it locally to your computer. And directly below that, we have the close-up image being fed into a video node with a prompt to get our character to speak. You can change the prompt to be whatever you'd like, but the default here is just generic information that the model makes up. >> Hi, I'm Alex. I really enjoy drawing cartoons in my free time, and my favorite dessert has to be chocolate fudge [music] cake. >> To get your character to say something specific, simply enter it into the prompt box along with some directions. >> Wow, this character workflow is awesome. Can you make me a robot to play with? You can continue to explore new characters or even add more nodes to customize this creation workflow even further. If you have any specific questions about workflows, make sure to check out the help articles. You can also head over to Discord to see how our creative community is using workflows. If you need to ask us a question directly, you can always use this button on your dashboard at any time. [music]
Original Description
Learn more about the Character Creator featured Workflow.
0:00 Intro
0:32 Explaining the Workflow nodes
1:39 New character pose example
2:20 Back to the main Workflow
4:07 Outro
—
runwayml.com
Watch on YouTube ↗
(saves to browser)
Sign in to unlock AI tutor explanation · ⚡30
Playlist
Uploads from Runway · Runway · 50 of 60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
▶
51
52
53
54
55
56
57
58
59
60
David Ha - Creative AI Conversations
Runway
Devi Parikh - Creative AI Conversations
Runway
Rebecca Fiebrink - Creative AI Conversations
Runway
Soumyadip Sengupta - Creative AI Conversations
Runway
Jia-Bin Huang - Creative AI Conversations
Runway
Something is coming... | Runway
Runway
Introducing Gen-2: Text to Video | Runway
Runway
Gen-2 Now Available | Runway
Runway
"Smile" by Ta-ku: Behind The Scenes | Runway Studios
Runway
A New Era of Creativity | Runway
Runway
In Conversation With: Julien Vallée | Runway Creative Partners Program
Runway
Introducing General World Models | Runway
Runway
Bias in AI and How to Fix It | Runway
Runway
Gen-3 Alpha: Available Now | Runway
Runway
Introducing Act-One | Runway
Runway
Getting Started with Act-One | Runway Academy
Runway
Behind the Scenes: Migration — Jeremy Higgins | Runway Studios
Runway
Using a Restyled First Frame with Video to Video | Runway Academy
Runway
Introducing Runway Gen-4 | Runway
Runway
The Herd | Gen-4 | Runway
Runway
Scimmia Vede | Gen-4 | Runway
Runway
The Retrieval | Gen-4 | Runway
Runway
The Lonely Little Flame | Gen-4 | Runway
Runway
NYC is a Zoo | Gen-4 | Runway
Runway
Gen-4 Image to Video | Runway Academy
Runway
Using 3D Assets with Gen-4 + References | Runway
Runway
Using 3D Assets in Gen-4 References | Runway Academy
Runway
Generating Environments with Gen-4 References | Runway
Runway
Runway This Week | July 7
Runway
Introducing Act-Two | Runway
Runway
Runway This Week | July 14
Runway
Expressive Character Performances | Runway Academy
Runway
Introducing Runway Aleph | A new way to edit, transform and generate video.
Runway
Runway This Week | August 2
Runway
How to Transform Videos with Runway Aleph | Runway Academy
Runway
Runway This Week | August 10
Runway
Runway This Week | August 17
Runway
Runway This Week | August 29
Runway
Runway This Week | September 6
Runway
Runway This Week | September 13
Runway
Using Workflows | Runway Academy
Runway
How to Build Custom Workflows | Runway Academy
Runway
Runway This Week | November 22
Runway
Introducing Gen-4.5 | Runway
Runway
Research Demo Day 2025 | Runway
Runway
Gen-4.5 Updates — Research Demo Day 2025 | Runway
Runway
GWM Real-time Worlds — Research Demo Day 2025 | Runway
Runway
GWM Avatars — Research Demo Day 2025 | Runway
Runway
GWM Robotics — Research Demo Day 2025 | Runway
Runway
Character Creator Workflow | Runway Academy
Runway
Move with Gen-4.5 | Runway
Runway
How to Use Motion Sketch | Runway
Runway
How to Turn an Image into an Entire Story with Story Panels Workflow | Runway
Runway
Previs your shoot with Storyboard Creator Workflow | Runway
Runway
All the best models. All inside Runway. | Runway
Runway ML
Morningstar: An AI short film | Runway
Runway ML
Introducing Runway Characters | Runway
Runway ML
Create Your Own Character in 5 Minutes | Runway
Runway
Build a Characters React App | Runway
Runway
Screen Sharing and Camera Feed with Runway Characters | Runway
Runway
More on: LLM Engineering
View skill →Related Reads
📰
📰
📰
📰
Corvorum OS 1.0 - Sistema Operativo Tecnomántico
Dev.to · Technomantus Corvi
Why Materials Scientists Are Still Copy-Pasting Data from PDFs in 2026 (And Why AI Changes…
Medium · AI
AI Automation Update: July 02
Dev.to AI
SpaceX has an AI device prototype, and it sure sounds phone-ish
TechCrunch AI
Chapters (5)
Intro
0:32
Explaining the Workflow nodes
1:39
New character pose example
2:20
Back to the main Workflow
4:07
Outro
🎓
Tutor Explanation
DeepCamp AI