<h1 id="does-everything-a-ponyxl-model-can-but-&quot;nearly&quot;-photorealistic!-e7zu6ys62">Does everything a PonyXL model can, but "nearly" photorealistic!</h1><p><span style="color:rgb(190, 75, 219)">V4 update, improved faces and lighting trained in.</span></p><p>Workflow I use to create the preview images <a rel="ugc" href="https://civarchive.com/models/646185/sdxlpony-face-and-upscale-civitai-metadata">sdxlpony-face-and-upscale-civitai-metadata</a></p><h3 id="what-it-does:-zhmjdxjed">What it does:</h3><p>If you're familiar with Pony models, they are very specific in their prompting and subject focus. I made this model to work the same way, but get closer to photorealistic. It uses the same prompting and LoRAs that other PonyXL models use.</p><p></p><h3 id="what-it-doesn't-do:-t1c5o1fvs">What it doesn't do:</h3><p>It's not a general model. Like other PonyXL models, it's more designed to work with booru tags and mostly does images with people in them. If you want to make general images, use a standard SDXL model.</p><p></p><h3 id="what's-coming:-n7vcxnbw3">What's coming:</h3><p>PonyXL is trained on images with "exaggerated" anatomy. So I've purposely kept it from getting too photorealistic for the moment due to uncanny valley issues. As I train it more, I'll bring it closer to photoreal. In the meantime, use prompts such as (realistic photograph, depth of field, bokeh, etc) to help get the photo look.</p><p></p><h3 id="how-to-use:-0poifahpa">How to use:</h3><p>I recommend using the standard SDXL VAE. CFG should range from 7 to 10, higher than other models. You can experiment with different samplers. I prefer DPM++ 2S a Karras for consistency and DPM++ 3M SDE Exponential to get wild. Prompting for quality should use the score system like other PonyXL models. (score_9, score_8_up, score_7_up) in the positive prompt and (score_3_up, score_4_up, score_5_up) in the negative prompt. Otherwise your images will be very plain.</p><p>Alternately, you can use the <a target="_blank" rel="ugc" href="https://civarchive.com/models/332646/pony-pdxl-negative-embeddings">Pony PDXL embeddings</a> that I created for much easier use.</p>

Everclear PNY by Zovya

everclearPNYByZovya_v2VAE.safetensors

EverClearPNY by zooya

V2+VAE

V1+VAE

<p>Does everything a PonyXL model can, but "nearly" photorealistic!</p><p><span style="color: rgb(190, 75, 219)">V3 update, increased face detail and better lighting trained in.</span></p><p>What it does:</p><p>If you're familiar with Pony models, they are very specific in their prompting and subject focus. I made this model to work the same way, but get closer to photorealistic. It uses the same prompting and LoRAs that other PonyXL models use.</p><p></p><p>What it doesn't do:</p><p>It's not a general model. Like other PonyXL models, it's more designed to work with booru tags and mostly does images with people in them. If you want to make general images, use a standard SDXL model.</p><p></p><p>What's coming:</p><p>PonyXL is trained on images with "exaggerated" anatomy. So I've purposely kept it from getting too photorealistic for the moment due to uncanny valley issues. As I train it more, I'll bring it closer to photoreal. In the meantime, use prompts such as (realistic photograph, depth of field, bokeh, etc) to help get the photo look.</p><p></p><p>How to use:</p><p>I recommend using the standard SDXL VAE. CFG should range from 7 to 10, higher than other models. You can experiment with different samplers. I prefer DPM++ 2S a Karras for consistency and DPM++ 3M SDE Exponential to get wild. Prompting for quality should use the score system like other PonyXL models. (score_9, score_8_up, score_7_up) in the positive prompt and (score_3_up, score_4_up, score_5_up) in the negative prompt. Otherwise your images will be very plain.</p><p>Alternately, you can use the <a target="_blank" rel="ugc" href="https://civitai.com/models/332646/pony-pdxl-negative-embeddings">Pony PDXL embeddings</a> that I created for much easier use.</p>

2.0+vae

<h1 id="does-everything-a-ponyxl-model-can-but-&quot;nearly&quot;-photorealistic!-e7zu6ys62">Does everything a PonyXL model can, but "nearly" photorealistic!</h1><p><span style="color:rgb(190, 75, 219)">V4 update, improved faces and lighting trained in.</span></p><p>Workflow I use to create the preview images <a rel="ugc" href="https://civitai.com/models/646185/sdxlpony-face-and-upscale-civitai-metadata">sdxlpony-face-and-upscale-civitai-metadata</a></p><h3 id="what-it-does:-zhmjdxjed">What it does:</h3><p>If you're familiar with Pony models, they are very specific in their prompting and subject focus. I made this model to work the same way, but get closer to photorealistic. It uses the same prompting and LoRAs that other PonyXL models use.</p><p></p><h3 id="what-it-doesn't-do:-t1c5o1fvs">What it doesn't do:</h3><p>It's not a general model. Like other PonyXL models, it's more designed to work with booru tags and mostly does images with people in them. If you want to make general images, use a standard SDXL model.</p><p></p><h3 id="what's-coming:-n7vcxnbw3">What's coming:</h3><p>PonyXL is trained on images with "exaggerated" anatomy. So I've purposely kept it from getting too photorealistic for the moment due to uncanny valley issues. As I train it more, I'll bring it closer to photoreal. In the meantime, use prompts such as (realistic photograph, depth of field, bokeh, etc) to help get the photo look.</p><p></p><h3 id="how-to-use:-0poifahpa">How to use:</h3><p>I recommend using the standard SDXL VAE. CFG should range from 7 to 10, higher than other models. You can experiment with different samplers. I prefer DPM++ 2S a Karras for consistency and DPM++ 3M SDE Exponential to get wild. Prompting for quality should use the score system like other PonyXL models. (score_9, score_8_up, score_7_up) in the positive prompt and (score_3_up, score_4_up, score_5_up) in the negative prompt. Otherwise your images will be very plain.</p><p>Alternately, you can use the <a target="_blank" rel="ugc" href="https://civitai.com/models/332646/pony-pdxl-negative-embeddings">Pony PDXL embeddings</a> that I created for much easier use.</p>

Files with SHA256: 4ce224f5736183dc1992cb9c1037178400f07cfc9a81c5dcfc1d6947b08eebfd