🪄 Provides easy&fast access to text2image / image2image / inpainting AI. (57 models across Stable Diffusion 1.x/2.x and finetuned models)
Fast AI image generation with the Dezgo API
This API provides you access to fast priority processing using the same model powering the web service Dezgo.com, with more features:
- fast generation times
- more features
- full access to all parameters
Enjoy the following features:
- text-to-image: create a brand new image from a text description
- image-to-image: transform an existing image according to a text description
- inpainting: transform a part of an existing image according to a text description
- upscaling : increase image resolution in a smart way
Dezgo.com uses the same API under the hood. If you want more (no rate-limiting, more steps count, bigger resolutions), then this API is for you!
We provide a cheap $2 plan to get you started.
ℹ️Please note that due to credit card abuse, we no longer offer pure pay-as-you go plans.
Tips for beginners
If you’re just getting started and want to get the most images out of your plan: stay with the standard resolution of 512x512, higher resolutions become quickly more expensive
With the $2 plan and default settings you can generate about 1,000 images.
Once you got your feet wet, you can always upgrade to a superior plan and use bigger resolutions and/or more sampling steps.
Credits
This API uses a credit system. Each call may consume more or less credits depending on the parameters you supply. This is directly related to the underlying compute resources consumed.
👉 With the BASIC plan for just $2 you can get 20,000 credits to get you started ($0.0001 / credit).
👉 The PRO, ULTRA and MEGA subscriptions provide you with a quota of credits at discounted rates (-10%, -15%, -20% respectively). This is handy if you expect to use the API more.
Example of request costs
Here are some examples of processing costs for images of different resolutions with the default settings.
(the dollar amounts represent the base credit cost, so do not take into account any eventual discounts)
/text2image, /image2image & /inpainting
| Resolution | Credits | $ |
|---|---|---|
| 512x512 | 19 | $0.0019 |
| 768x768 | 87 | $0.0087 |
| 1024x1024 | 181 | $0.0181 |
If you use the upscale parameter, then upscale costs are added to the above costs.
This is a great way to save money as upscaling is cheaper than generating directly at higher resolutions.
Example, say you want to generate an image with an output resolution of 1024x1024:
| Example | Parameters | Credits | $ |
|---|---|---|---|
| native 1024px | width=1024 height=1024 |
181 | $0.0181 |
| upscaled from 512px to 1024px | width=512 height=512 upscale=2 |
31 (19+12) | $0.0031 |
/upscale
| Resolution (of the input image) | Credits | $ |
|---|---|---|
| 320x320 | 5 | $0.0005 |
| 512x512 | 12 | $0.0012 |
| 768x768 | 29 | $0.0029 |
| 1024x1024 | 51 | $0.0051 |
Request cost calculation
The exact formulas for each method are as follows:
/text2image, /image2image & /inpainting
pixels = max(width * height, 512 * 512)
credits_cost = (pixels * 6.849e-006 - 1.133) * max(steps, 10)
If the upscale parameter is used, then the costs from /upscale are added to the above costs.
/upscale
pixels = max(width * height, 320 * 320)
credits_cost = pixels * 4.9278E-05
Credits information in HTTP Response Headers
You can analyze your quota usage with the following HTTP headers:
x-credits-cost: credits charged for the current requestx-ratelimit-credits-remaining: credits currently remaining in your quotax-ratelimit-credits-limit: the initial amount of credits allowed by your quota.x-ratelimit-credits-reset: seconds remaining until the next quota reset
Custom plans
If you expect to send a very high volume of requests, contact us, we can offer a custom private plan that fits your needs.