Image First: Why Making a Still Before the Video Saves Credits
Text to video asks the model to invent everything: subject, setting, lighting and motion, all at once. The image-first workflow splits that into two steps. You make and perfect a still image, then animate it. You see exactly what you are getting before you spend on motion, and it often costs less overall.
- Make a still in Text to Image (1 credit), refine it, then animate it in Image to Video.
- You control the look before paying for motion.
- It also keeps a character or product consistent across clips.
- Use text to video when you want the model to surprise you.
The Two Routes
| Text to video | Image first | |
|---|---|---|
| What the model decides | Everything | Only the motion |
| Control over the look | Low | High |
| Cost per attempt (cheapest model) | 2 credits (Kling 2.5 Turbo Pro) | 1 for the image + 1 for the clip = 2 |
| Retries when the look is wrong | Re-run the whole clip | Re-make the 1-credit image |
| Best for | Exploration, atmosphere, surprises | Specific looks, products, characters, series |
The Workflow
An image plus a clip is 2 credits.
Why It Often Costs Less
Suppose you want a specific scene and expect to try four versions before you like the look. With text to video on Kling 2.5 Turbo Pro that is 4 x 2 = 8 credits, and each attempt changes motion and look together. With image first, you try four stills for 4 credits, pick the best, and animate it once or twice for 1 to 2 credits, a total of 5 to 6. You also know what the clip will look like before it moves.
These are illustrations. Your own numbers will differ, but the pattern holds: looks are cheap to iterate as images, and expensive as video.
When to Use Text to Video Instead
- Atmospheric footage where any good result will do.
- Models with built-in sound in the Premium and Next-Gen tabs, which also accept text.
- Long clips in Next-Gen, where a start image is optional.
- When you want to be surprised.
See text to video vs image to video for a fuller comparison.
Image-First for Series and Brands
A reference image gives you a fixed starting point for every shot. For characters, edit the same image for new poses. For products, shoot or generate one clean master image and make several clips from it. See keeping characters consistent and storyboarding an AI video.
Prompt Tips for the Animation Step
- Do not repeat what the image already shows. Describe the motion.
- Say what should stay the same.
- Keep the camera move simple.
- Use the same aspect ratio as the image to avoid cropping.
More wording in our prompting guide.
Frequently Asked Questions
What is the image-first workflow?
Make and refine a still image in Text to Image, then animate it in Image to Video, instead of generating the video directly from text.
Is image first cheaper than text to video?
Often. Stills cost 1 credit each, so you can iterate on the look cheaply, then animate the best one. A text-to-video attempt on Kling 2.5 Turbo Pro costs 2 credits.
How do I send an image to video?
Use the Send to Video button on the Text to Image result, which carries the image into the Image to Video tab.
When should I use text to video instead?
For exploration, atmosphere, long Next-Gen clips and models with built-in sound, or when you want the model to surprise you.
Does image first help with consistency?
Yes. A fixed reference image gives every clip the same starting point.
Related Articles
- Text to Video vs Image to Video: Which to Use (2026)
- How to Animate Any Photo with AI in 2026 (Step by Step)
- How to Keep a Character Consistent Across AI Video Clips
- How to Storyboard an AI Video Before You Spend a Credit
- Image to Video AI: Complete Guide for 2026 (All Models Compared)
Try It on Your Own Project
Buy a small amount of credits with no subscription, or subscribe from $47/mo. New customers get 20% off their first card payment.
⚡ Get Started →20% off first card payment automatically · Plans from $47/mo · Or pay-per-use, no subscription