replicate / image
qwen-image-lora-trainer-legacy
Fine-tunable Qwen Image model with exceptional composition abilities - train custom LoRAs for any style or subject
qwen/qwen-image-lora-trainer-legacy
Source checked: 2026-09-13 · Provider endpoint, not an independently benchmarked model.
Pricing, with its conditions.
These are source observations, not interchangeable quotes. Token, compute-time, output-duration and per-request prices cannot be compared by the number alone. Input charges and optional features may add to the total.
| Source rate | Billing unit | Conditions | Evidence |
|---|---|---|---|
| $1.525 USD | second of run time | or around 655 seconds for $1 | Public structured source |
Supported settings from the source
| width | Custom width in pixels. Provide both width and height for custom dimensions (overrides aspect_ratio/image_size). · Min: 512 · Max: 2048 |
|---|---|
| height | Custom height in pixels. Provide both width and height for custom dimensions (overrides aspect_ratio/image_size). · Min: 512 · Max: 2048 |
| aspect_ratio | Aspect ratio for the generated image. Ignored if width and height are both provided. · Options: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3 · Default: 16:9 |
| output_format | Format of the output images · Options: webp, jpg, png · Default: webp |
| output_quality | Quality when saving images (0-100, higher is better, 100 = lossless) · Min: 0 · Max: 100 · Default: 80 |
Unlisted settings, region availability, deposits, taxes and commercial terms remain unverified. Confirm the exact endpoint before purchase.