LTX-2 IC-LoRA Trainer lets you train custom In-Context LoRA models for video-to-video transformations, including depth/pose adapters, video restoration, and style transfer. Upload a ZIP file containing paired videos to start. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Idle
Your request will cost $0.75 per run.
For $10 you can run this model approximately 13 times.
README
LTX-2 19B IC-LoRA Trainer
LTX-2 19B IC-LoRA Trainer is a high-performance custom model training service for the LTX-2 19B video generation model. It allows you to train lightweight LoRA (Low-Rank Adaptation) adapters for personalized styles, characters, and concepts — bringing your custom visuals into AI-generated videos with synchronized audio.
Why Choose This?
-
Video LoRA training Train custom adapters specifically optimized for LTX-2's 19B DiT video architecture.
-
Character consistency Create LoRAs that maintain character identity across generated video clips.
-
Style personalization Capture specific artistic styles, brand aesthetics, or visual themes for video content.
-
Seamless integration Trained LoRAs work directly with LTX-2 Text-to-Video LoRA and Image-to-Video LoRA models.
Training Process
-
Data Upload Prepare and upload a ZIP file containing your training videos. Include 3-10 high-quality, diverse video clips for best results.
-
Configure Trigger Word Set a unique trigger word (e.g., "p3r5on") that will activate your trained style or character in prompts.
-
Adjust Training Parameters
- steps — Total training iterations (default: 500)
- learning_rate — Training speed (default: 0.0002)
- lora_rank — Adapter capacity (default: 32)
- LoRA Training The system runs a tailored LoRA optimization loop:
- Freezes the base model weights
- Trains only the low-rank adapter layers
- Applies LTX-2 optimized settings for video generation
- Model Export After training completes, you receive a LoRA adapter file (.safetensors) compatible with:
Parameters
| Parameter | Default | Description |
|---|---|---|
| data | — | ZIP file containing training videos (required) |
| trigger_word | — | Unique word to activate your trained concept |
| steps | 500 | Total training iterations |
| learning_rate | 0.0002 | Training speed (lower = more stable, higher = faster) |
| lora_rank | 32 | Adapter capacity (higher = more detail, larger file) |
Pricing
| Training Steps | Price (USD) |
|---|---|
| 100 | $0.75 |
| 500 | $3.75 |
| 1,000 | $7.50 |
| 2,000 | $15.00 |
Billing Rules
- Base price: $0.75 per 100 steps
- Total cost = $0.75 × (steps / 100)
- Billed proportionally to the total number of steps in your job
Best Use Cases
- Character LoRAs — Train on character videos to maintain identity across video generations.
- Brand Styles — Create custom visual styles for consistent marketing video content.
- Art Styles — Capture specific artistic aesthetics for creative video projects.
- Motion Patterns — Train on specific movement styles or animation techniques.
Pro Tips
- Use 3-10 high-quality, diverse video clips of your subject for best results.
- Choose a unique trigger word that won't conflict with common words.
- Higher lora_rank (32-64) captures more detail but increases training time and file size.
- Lower learning_rate is more stable but requires more steps.
- Start with default settings, then adjust if needed.
Try More Trainers
- Qwen Image 2512 LoRA Trainer — Train LoRAs for Qwen Image text-to-image model.
- Z-Image LoRA Trainer — Train LoRAs for Z-Image models with Turbo-compatible optimization.
- Flux Dev LoRA Trainer — Train LoRAs for Flux Dev image generation model.
Guidance
Notes
- Higher parameter values (steps, lora_rank) will increase training time.
- Training time scales with the number of video clips and total steps configured.
- For faster iterations, start with lower settings and increase gradually.