Wan 2.2 T2V 720p LoRA API Documentation

Wan 2.2 T2V 720p LoRA API Documentation

Playground

Wan 2.2 T2V 720p with custom LoRA support turns text prompts into 720p AI videos and enables unlimited video generation. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Wan 2.2 Text-to-Video 720p LoRA is a powerful text-to-video generation model that creates stunning 720p HD videos from text descriptions. With advanced LoRA support including high-noise and low-noise options, apply custom styles, artistic effects, or consistent character appearances to create unique video content.

Why It Stands Out

• HD 720p output: Generate crisp videos in landscape (1280×720) or portrait (720×1280) formats.
• Advanced LoRA support: Three types of LoRA inputs for precise style control.
• Prompt Enhancer: Built-in AI-powered prompt optimization for better results.
• Negative prompt support: Exclude unwanted elements for cleaner outputs.
• Flexible duration: Choose between 5 or 8 second video lengths.
• Reproducibility: Use the seed parameter to recreate exact results.

Parameters

How to Use

1. Write a prompt describing the scene, action, and style you want. Use the Prompt Enhancer for AI-assisted optimization.
2. Add a negative prompt (optional) — specify elements to exclude.
3. Select size — choose landscape (1280×720) or portrait (720×1280).
4. Set duration — choose 5 or 8 seconds.
5. Add LoRAs (optional) — apply standard, high-noise, or low-noise LoRAs.
6. Click Run and wait for your video to generate.
7. Preview and download the result.

Understanding LoRA Types

• Standard LoRAs: Applied throughout the generation process.
• High-Noise LoRAs: Applied during early denoising steps for structural/compositional effects.
• Low-Noise LoRAs: Applied during later denoising steps for fine detail and style refinement.

Combining different LoRA types gives you precise control over the final output.

Best Use Cases

• Creative Animation — Apply unique visual styles and effects.
• Social Media Content — Create platform-optimized videos for TikTok, Reels, and Shorts.
• Marketing & Advertising — Produce stylized promotional videos.
• Artistic Projects — Generate videos with specific aesthetic styles.
• Character Consistency — Maintain character appearance across multiple videos.

Pricing

Pro Tips for Best Quality

• Be detailed in your prompt — describe subject, action, environment, lighting, and mood.
• Use negative prompts to reduce artifacts like blur, distortion, or unwanted motion.
• Experiment with different LoRA combinations for unique effects.
• Use high-noise LoRAs for overall style, low-noise LoRAs for detail refinement.
• Choose portrait (720×1280) for mobile-first platforms like TikTok.
• Fix the seed when iterating to compare different LoRA combinations.

Notes

• Ensure uploaded LoRA paths are correct and accessible.
• Processing time varies based on duration and current queue load.
• Please ensure your prompts comply with content guidelines.

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",
  "size": "1280*720",
  "duration": 5
}
JSON
)

# 1. Submit the prediction once. Never retry this request automatically:
# a retry after an unclear failure can create a second, billed task.
CHECK='The task may already exist. Check History or the predictions list before submitting again.'
if ! SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  --connect-timeout 10 --max-time 60 \
  -X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/wan-2.2/t2v-720p-lora" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}"); then
  printf 'Submit failed. %s\n' "${CHECK}" >&2
  exit 1
fi

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 "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id. %s\n' "${CHECK}" >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 2. Poll until the prediction reaches a final status.
BODY_FILE=$(mktemp)
trap 'rm -f "${BODY_FILE}"' EXIT
while true; do
  # Transient errors (429, 5xx) are retried, honoring Retry-After.
  HTTP_CODE=$(curl --silent --show-error --fail-with-body \
    --connect-timeout 10 --max-time 60 --retry 5 --retry-delay 2 \
    --output "${BODY_FILE}" --write-out '%{http_code}' \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}" || true)
  case "${HTTP_CODE}" in
    2??) ;;
    000|429|5??)  # dropped connection or temporary error: keep polling
      printf 'Task %s: HTTP %s, polling again in 10s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      sleep 10; continue ;;
    *) printf 'Task %s: HTTP %s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      cat "${BODY_FILE}" >&2; exit 1 ;;
  esac
  RESULT=$(jq 'if type == "object" and has("data") then .data else . end' "${BODY_FILE}")
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout|deleted)
      printf 'Task %s ended with status %s\n' "${PREDICTION_ID}" "${STATUS}" >&2
      printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    *) sleep 2 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Every model also accepts the global request options enable_sync_mode and enable_base64_output; see their pages for details.

Response Parameters

Result Request Parameters

Result Response Parameters