APIMaster.ai

Vidu Image Q2 · API reference

Submit text-to-image and reference-image requests, poll tasks, and download Vidu Image Q2 results.

Create a task with POST /v1/wand/vidu-image/generation and query it with GET /v1/wand/vidu-image/tasks/{task_id}. Send an APIMaster API key in the Authorization: Bearer header with both requests. The POST body is JSON.

Field Type Description
model string Required. Use vidu-image-q2.
prompt string Required. Up to 2,000 characters. Omit reference images for text-to-image.
images string[] Optional. Up to seven public URLs or data URIs. PNG, JPEG/JPG, and WebP are supported. Each image must be at least 128 × 128 pixels, with aspect ratio from 1:4 to 4:1 and size up to 50 MB. The whole request must stay within 20 MB.
aspect_ratio string 16:9 by default. Options are 16:9, 9:16, 1:1, 3:4, 4:3, 21:9, 2:3, 3:2, and auto. The auto setting follows the first reference image.
resolution string 1080p by default. Also supports 2K and 4K. Values are case-sensitive.
seed integer Omit or set to 0 for a random seed.
curl -X POST "https://apimaster.ai/v1/wand/vidu-image/generation" \
  -H "Authorization: Bearer $APIMASTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"vidu-image-q2","prompt":"A blue circle on a white background","resolution":"1080p"}'

Use the returned task_id and the same account to poll every 3–5 seconds. Possible states are created, queueing, processing, success, and failed. On success, download the files under creations[].url within 12 hours. Use request_id to trace failures such as invalid images, content review, insufficient credit, rate limits, or missing tasks.

Polling does not resubmit the task or charge it again. callback_url is unsupported. To recover a lost submission, send a stable Idempotency-Key of 8–128 visible ASCII characters and follow the submission recovery guide. Reusing the key with the same parameters returns a 202 receipt. Different parameters with the same key return 409.

tokenhub_usage.total_tokens reports the billable usage, which depends on resolution and how many references were used. See the Vidu Image Q2 model card for the current APIMaster price.