Workflows:

Eros Workflows - Mr.Xin

EasyFlow

10Eros V1 (coming soon):

• This is a minor finetune-level adjustment of LTX 2.3 using data from the Sulphur2 project. It is entirely I2V focused. It will work T2V, but you're on your own.

• I didn't train it but had access to all steps and implementation and have permission to share it. This is merged with a custom script. This exact mix seemed the best for all-around use, but it's far from perfect. Further updates will most likely be enhanced with my own personal lora training to cover weaknesses.

• This is not official or in any way affiliated with Lightricks.

• If you want to host this commercially you're better off contacting me than just running it.

Quality is entirely prompt, image, and workflow dependent. Aside for systemic issues with LTX: motion smear, artifacts, and identity loss, it still behaves like 2.3 requiring extensive prompt refinement, extra sampling, and resolution targeting for best results.

This is not simple AIO lora merge, it uses no on-site loras; any 2.3 loras can be stacked on top of it and if you need them use them. It's a layer-scaled integration of 768-ranked data with a vibe-coded custom script. It fills in the nsfw and explicit motion gaps in the base 2.3 model without fully intruding on the 22b dev model's functionality and audio, at least an attempt to do so. New anatomy is depictable and introducible but subject to quality loss at low resolution and sampling steps and definitely not always fully accurate. All general nsfw and explicit ideas are within reach with proper prompting and sampling, some full form changes or scene changes may require additional conditioned frames for best results.

Common issues:

1. Motion smear during fast motion - still working on a fix. LTX team needs to make addressing this their top priority. I think it's somewhat related to the ever-problematic distilled loras (they're garbage imho). If you have a perfect motion output but it's ruined by this you can interpolate it and run it back into the model as a 24fps video input and re-render it to reduce it. It is especially bad in this version due to the possibility of actual fast and lively motion being reached with prompt.

2. Likeness change - Prompt/Img_compression dependent. The only conditioning you get with a single i2v input is the first frame afai cant tell. If likeness matters you need at least one middle frame or end frame conditioning to keep it. There will probably be an IC Lora or node that could be made to enhance the facial conditioning.

3. Artifacts, extra limbs, body contortion - Under sampling, bad prompt, img_compression too low/high, video too long/short for prompt. For instance, don't just prompt "ass" when all of the action is her facing frontal to the camera and you have no detailed prompting for her to turn around and show it.

4. Vertical upscale artifacts - this was supposedly fixed, it is not. Extreme vertical aspect ratios that upscale to 1280+ will have body morphing. Crop the input more square.

5. Bad quality/appearance, just terrible output - adjust distilled lora strength, prompt, image. The model can make a lot of images and ideas work but not everything.

This probably means nothing legally, but I just want to make it clear that you waive any right to blame or use the host site, me, or sulphur in your defense if you download and misuse this. It's ethical artistic intended use is clearly stated:

By downloading, accessing, or using the 10Eros Model (the “Model”), you (“User”) agree to the following terms:

1. Permitted Use Only The Model may only be used with inputs of fictional or non-real characters. Any use involving real persons, celebrities, public figures, or identifiable individuals is strictly prohibited (“Out-of-Scope Use”).

2. Disclaimer of Liability TenStrip (the creator of the Model), together with its affiliates (CivitAI.com / CivitAI.red, Lightricks Ltd., and Sulphur Team), collectively and individually disclaim all liability for any direct, indirect, incidental, consequential, special, or punitive damages arising from your use or misuse of the Model.

This includes, without limitation, any claims related to copyright infringement, right of publicity violations, defamation, privacy violations, deepfakes, illegal activities, or any harm caused by outputs generated by the Model.

3. Out-of-Scope Uses Forbidden All Out-of-Scope Uses are explicitly forbidden. The creator and affiliates bear no responsibility for any outputs generated from prohibited inputs. Your actions with this generative Model are solely your own.

4. Assumption of Risk You assume all risks associated with the use of the Model. Any use outside the permitted fictional/non-real character inputs further releases TenStrip and all affiliates from any and all liability.

5. Indemnification You agree to indemnify, defend, and hold harmless TenStrip and its affiliates from any and all claims, losses, liabilities, damages, and expenses (including reasonable attorneys’ fees) arising from your use or misuse of the Model.

6. No Warranties The Model is provided “AS IS” with no warranties of any kind, express or implied.

This agreement is governed by the laws of the United States. If you do not agree to these terms, do not download or use the Model.