Qwen Image Edit LoRA API Documentation
Playground
Try it on WaveSpeedAI!Qwen-Image-Edit LoRA (20B) enables bilingual Chinese/English image-to-image editing with style preservation and semantic and appearance edits. Ready-to-use REST API, best performance, no coldstarts, affordable pricing.
Features
Qwen-Image-Edit with LoRA is a 20B MMDiT-based next-gen image editing model, Built on 20B Qwen-Image, it brings precise bilingual text editing (Chinese & English) while preserving style, and supports both semantic and appearance-level editing.
Key Features
-
Precise bilingual text editing: Directly add, delete, or modify text in Chinese or English, while preserving font, size, kerning, and style.
-
LoRA integration: Import up to 3 external LoRA weights (.safetensors), each with its own blending scale, for tailored effects.
-
Style preservation: Maintains palette, lighting, and overall artistic intent even under substantial edits.
-
SOTA benchmark results: Achieves state-of-the-art performance across multiple public image editing benchmarks.
Limits and Performance
- Max resolution per job: up to 1536 × 1536 pixels
- Max LoRAs: 3 per job (with individual scaling controls)
- Output formats: JPEG / PNG / WEBP
- Processing speed: ~6–12 seconds per image
- Input: Requires image + prompt (can include editing instructions and/or text edits)
Pricing
- $0.025 per image
- Each generated image is billed individually.
How to Use
- Upload or paste a link to your source image.
- Write a prompt describing desired edits (appearance or semantic).
- (Optional) Add up to 3 LoRAs:
- Provide LoRA path/URL.
- Adjust scale for each (0.1–1.0 recommended).
- Adjust size (width & height, up to 1536×1536).
- (Optional) Add a seed for reproducibility.
- Run the job → preview results → refine with prompt or LoRA scaling.
Pro tips for best results
- Use appearance editing for clean local changes (e.g., shirt color).
- Use semantic editing for creative/global changes (e.g., pose, style transfer).
- For text edits, clearly specify text content + style in the prompt.
- Combine LoRAs for hybrid results, but keep scale balanced (too high may distort).
- Lock the seed when testing multiple LoRAs to compare effects consistently.
Note
- If you did not upload the image locally, please ensure that the image URL is accessible! A successfully accessible image will display a preview in the interface.
Reference
Authentication
For authentication details, please refer to the Authentication Guide.
API Endpoints
Submit Task & Query Result
set -euo pipefail
export WAVESPEED_API_KEY="your-api-key"
REQUEST_BODY=$(cat <<'JSON'
{
"prompt": "A cinematic ocean wave at sunrise, highly detailed",
"image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
"seed": -1,
"output_format": "jpeg"
}
JSON
)
# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
-X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/qwen-image/edit-lora" \
-H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
-H "Content-Type: application/json" \
-d "${REQUEST_BODY}")
TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
printf 'Submission response did not contain a prediction id
' >&2
exit 1
fi
RESULT_URL=$(printf '%s' "${TASK}" | jq -r '.urls.get // empty')
if [ -z "${RESULT_URL}" ]; then RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"; fi
# 2. Poll until the prediction finishes.
while true; do
RESPONSE=$(curl --silent --show-error --fail-with-body \
"${RESULT_URL}" \
-H "Authorization: Bearer ${WAVESPEED_API_KEY}")
RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')
case "${STATUS}" in
completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
failed|cancelled|timeout) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
created|processing) sleep 2 ;;
*) printf 'Unexpected status: %s
' "${STATUS}" >&2; exit 1 ;;
esac
doneParameters
Task Submission Parameters
Request Parameters
| Parameter | Type | Required | Default | Range | Description |
|---|---|---|---|---|---|
| prompt | string | Yes | - | The positive prompt for the generation. | |
| image | string | Yes | - | The image to generate an image from. | |
| loras | array<object> | No | 0 ~ 3 items | List of LoRAs to apply (max 3). | |
| seed | integer | No | -1 | - | The random seed to use for the generation. -1 means a random seed will be used. |
| output_format | string | No | jpeg | jpeg, png, webp | The format of the output image. |
| enable_base64_output | boolean | No | false | - | If set to `true`, the prediction's `output` strings are returned as **naked base64** (no `data:<mime>;base64,` prefix). When `false` (default), outputs are returned as URLs pointing to our CDN. |
| enable_sync_mode | boolean | No | false | - | If set to `true`, the request attempts to wait for the generated result and return outputs in the same response. If the result is not ready within the sync wait window, the API can return a timeout body while the task continues processing. This option is only available via the API and is supported only by some models. |
Response Parameters
| Parameter | Type | Description |
|---|---|---|
| code | integer | HTTP status code (e.g., 200 for success) |
| message | string | Status message (e.g., “success”) |
| data.id | string | Unique identifier for the prediction, Task Id |
| data.model | string | Model ID used for the prediction |
| data.outputs | array | Output values, usually URL strings; some models return text strings or structured result objects (empty when status is not completed) |
| data.urls | object | Object containing related API endpoints |
| data.urls.get | string | URL to retrieve the prediction result |
| data.status | string | Status of the task: created, processing, completed, or failed |
| data.created_at | string | ISO timestamp of when the request was created (e.g., “2023-04-01T12:34:56.789Z”) |
| data.error | string | Error message (empty if no error occurred) |
| data.timings | object | Object containing timing details |
| data.timings.inference | integer | Inference time in milliseconds |
Result Request Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
| id | string | Yes | - | Task ID |
Result Response Parameters
| Parameter | Type | Description |
|---|---|---|
| code | integer | HTTP status code (e.g., 200 for success) |
| message | string | Status message (e.g., “success”) |
| data | object | The prediction data object containing all details |
| data.id | string | Unique identifier for the prediction |
| data.model | string | Model ID used for the prediction |
| data.outputs | array<string | object> | Array of generated outputs (empty when status is not completed). Items are usually URL strings, but may be text strings or structured result objects, depending on the model. |
| data.urls | object | Object containing related API endpoints |
| data.urls.get | string | URL to poll for the prediction result |
| data.status | string | Status: created, processing, completed, or failed |
| data.created_at | string | ISO timestamp of when the request was created |
| data.error | string | Error message (empty if no error occurred) |
| data.timings | object | Object containing timing details |
| data.timings.inference | integer | Inference time in milliseconds |