This is going to be a quick rundown on how you can save up on Buzz. First and foremost, it's gonna save you a lot of headaches in the long run to just bite the bullet and buy 5,000 Buzz. You'll see why.

# Gaining Buzz Daily.

Doing your daily quests will earn you up to **225 Buzz **each day. Assuming you don't spend it on generations or other Buzz costing things, you can expect to earn **1575 Buzz** a week. In a month, that's roughly 6000 Buzz.

# Gaining Buzz Monthly

You can gain these passively and for three of them, you can knock them out in one move by using a Colab to create LoRAs and then using another site to create images for it (or use your computer if you have one equipped to handle it) before uploading both to Civit.

Assuming you get all these quotas filled up, you're looking at **51,200 Buzz** a month. Adding that together together and you'd get about **57,285** Buzz per month.

# Image Generation Cost

This is the meat and potatoes of the image.

The major cost contributor is Image Steps. What determines the price is the Model and Sampler, and what multiplies it is the Quantity. A small modifier is the size of the image, but it's not as much.

SD 1.5 is by far the cheapest as its our guideline to see the most steps it takes to retain a 1 Buzz cost. My findings are that:

1 Image, Euler a, Square = 37 Steps

1 Image, DPM ++ 2M Karras, Square = 25 Steps

1 Image, Euler a, Landscape= 25 Steps

2 Images, Euler a, Square = 25 Steps

2 Images, DPM ++ 2M Karras, Square = 12 Steps

2 Images, Euler a, Landscape = 12 Steps

Unfortunately, you cannot go lower than 10 steps so there is no way you can have 2 images with DPM ++ 2M Karras in a landscape/portrait setting.

SDXLs and Pony Diffusions instantly bump the price up and play by different rules. For one, no 1 Buzz images. Instead, the initial price is *2* Buzz and goes up every 9-10 steps in Euler A, with this applying earlier on DPM. Surprisingly, Landscape/Portrait settings make it so that you can take an extra step before the Buzz price hikes up.

Also, I may have made a mistake in that it doesn't multiply per image but rather just changes price in a percentage base because four images at 50 Steps costs 21 Buzz and not 2*4*?

# Breaking It All Down

Now, you probably want to know... What exactly does all this mean? The main take aways are that SD 1.5 is the cheapest Model, but if you want to do the more expensive SDXL/Pony LoRAs, then eye the number of steps and judge how many Buzz you're spending.

But, to make calculations easier, I've taken a look at all possible configurations of CFG, Samples, and Steps, and then calculated just how many images per day you can make just 50 Buzz so that you can make a net positive of 175 Buzz each day. This is assuming you're using 4 images in Portrait.

**SD 1.5, CFG 4, Euler, 14 Steps:**16 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 7, Euler, 20 Steps:**12 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 10, Euler, 26 Steps:**10 sets of images.**SD 1.5, CFG 4, Euler, 24 Steps:**12 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 7, Euler, 30 Steps:**10 sets of images.**SD 1.5, CFG 10, Euler, 36 Steps:**8 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 4, Euler, 34 Steps:**8 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 7, Euler, 40 Steps:**7 sets of images, with 1 Buzz remaining.**SD 1.5, CFG 10, Euler, 46 Steps (or 50 Steps):**6 sets of images, with 2 Buzz remaining.

**SD 1.5, CFG 4, DPM, 14 Steps:**12 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 7, DPM, 20 Steps:**10 sets of images.**SD 1.5, CFG 10, DPM, 26 Steps:**7 sets of images, with 1 Buzz remaining.**SD 1.5, CFG 4, DPM, 24 Steps:**8 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 7, DPM, 30 Steps:**6 sets of images, with 2 Buzz remaining.**SD 1.5, CFG 10, DPM, 36 Steps:**5 sets of images, with 5 Buzz remaining.**SD 1.5, CFG 4, DPM, 34 Steps:**5 sets of images, with 5 Buzz remaining.**SD 1.5, CFG 7, DPM, 40 Steps:**5 sets of images.**SD 1.5, CFG 10, DPM, 46 Steps (or 50 Steps):**4 sets of images, with 2 Buzz remaining.

**SDXL/Pony, CFG 4, Euler, 14 Steps:**6 sets of images, with 2 Buzz remaining.**SDXL/Pony, CFG 7, Euler, 20 Steps:**5 sets of images, with 5 Buzz remaining.**SDXL/Pony, CFG 10, Euler, 26 Steps:**4 sets of images, with 6 Buzz remaining.**SDXL/Pony, CFG 4, Euler, 24 Steps:**5 sets of images.**SDXL/Pony, CFG 7, Euler, 30 Steps:**3 sets of images, with 11 Buzz remaining.**SDXL/Pony, CFG 10, Euler, 36 Steps:**3 sets of images, with 5 Buzz remaining.**SDXL/Pony, CFG 4, Euler, 34 Steps:**3 sets of images, with 8 Buzz remaining.**SDXL/Pony, CFG 7, Euler, 40 Steps:**2 sets of images, with 16 Buzz remaining.**SDXL/Pony, CFG 10, Euler, 46 Steps:**2 sets of images, with 12 Buzz remaining.**SDXL/Pony, CFG 10, Euler, 50 Steps:**2 sets of images, with 8 Buzz remaining.

**SDXL/Pony, CFG 4, DPM, 14 Steps:**5 sets of images, with 5 Buzz remaining.**SDXL/Pony, CFG 7, DPM, 20 Steps:**3 sets of images, with 11 Buzz remaining.**SDXL/Pony, CFG 10, DPM, 26 Steps:**2 sets of images, with 16 Buzz remaining.**SDXL/Pony, CFG 4, DPM, 24 Steps:**3 sets of images, with 5 Buzz remaining.**SDXL/Pony, CFG 7, DPM, 30 Steps:**2 sets of images, with 12 Buzz remaining.**SDXL/Pony, CFG 10, DPM, 36 Steps:**2 sets of images, with 4 Buzz remaining.**SDXL/Pony, CFG 4, DPM, 34 Steps:**2 sets of images, with 8 Buzz remaining.**SDXL/Pony, CFG 7, DPM, 40 Steps:**2 sets of images**SDXL/Pony, CFG 10, DPM, 46 Steps:**1 set of images, with 21 Buzz remaining.**SDXL/Pony, CFG 10, DPM, 50 Steps:**1 sets of images, with 19 Buzz remaining.

Averaging this out, this means that on average, SD 1.5 with Euler A wins out because with 50 Buzz, you could generate an average of 40 images in a single day...

At least, before you factor in the remaining Buzz. Suppose you leave those remaining Buzz be and they add to the 50 Buzz... Suddenly, the story becomes different, as the DPM Pony stats will have bolstered the Buzz count by nearly half of what you would have been using.

Or even suppose you use that Buzz *now* to eek out a few more images. For whatever reason, you can't make a set of two images with 21 Buzz but you *can* make two one-image sets for 8 Buzz, then have 5 Buzz remaining. Though, this only adds about one or two extra images to your averages.

The *real* kicker is adding the extra 50 Buzz on top. Let's take our most expensive options, Max Scale Pony DPM. Adding 50 Buzz onto your 21 remaining Buzz lets you make *two* sets of images with 13 remaining Buzz instead of 1. Having 100 outright triples the output with three sets of images a day. And this is at the max, most expensive option.

# Yes, but what does this all mean? (AKA: TL;DR)

**Be frugal with your Buzz spending**if you want to coast by with free Buzz gains. This may mean skipping generation days and only spending it browsing images and reacting.If possible, use the lowest Steps, Euler A, and the lowest quantity of images.

If you can, create LoRAs or images and contribute to the site's growth. You'll not only earn Buzz that way, but you'll also earn a name for yourself and may even get a chance to earn more Buzz down the line.

Participate in Bounties if possible: I know this is weird and last minute to put into this, but sometimes people will pay good Buzz just for people to generate images for their LoRAs. If you have a nack for LoRA creation or Dataset creation, even better.

If you truly want to generate endlessly, support CivitAI by buying 5,000 Buzz. Using the Balanced baselines of a Pony Model, you'd average out to about 384-555 sets of four images if you spend it all in one place. You'd most likely either find the image you're looking for or get cut off. If you pick a Monthly subscription of 10 dollars, not only do you get 10,000 Buzz to play with (which means the amount gets

*doubled*to 769-1,111 sets), but you earn a 50% increase of Buzz rewards from your Quest, which means your Daily and Monthly quotas jump up to**337****Buzz and a full 76,800 Buzz**respectively.

And that, I believe, would make it worth it in the end.