MiniMax H3 Open Weights ControlNet Union generates a new video that follows the motion and composition of a source video. Pose, depth, edges, lines, scribble or grayscale structure is extracted from the source automatically and guides the output, optionally with reference images for the subject or style, with native stereo audio generated in the same pass. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
Idle
$0.15per run·~66 / $10
The man from <Picture 1> - same face, curly blond hair, glasses, grey-green jacket and baggy trousers, white sneakers - dances on a neon-lit city street at night, full-body shot. Audio: upbeat hip-hop beat.
MiniMax H3 ControlNet Union Open Weights generates a new video that follows the motion and composition of a source video. The control signal - pose, depth, edges, straight lines, scribble or grayscale structure - is extracted from the source automatically, so you only upload a normal video, describe the result, and optionally add reference images for the subject or style. Native stereo audio is generated in the same pass, or the source soundtrack can be kept.
Seven control types in one model
pose, depth, canny, soft_edge, lines, scribble and gray, all extracted automatically from your video.
Character animation
Combine pose with a reference image to make your character perform the motion of the source video.
Restyle while keeping structure
Use edges, lines or depth to re-render a clip in a new style, setting or look while keeping its layout and camera movement.
Recolor and relight
Use gray to keep the brightness structure and change colors, lighting or grading from the prompt.
Native audio or preserved audio
Generate new audio, or set generate_audio to false to keep the source video's soundtrack.
| Parameter | Required | Description |
|---|---|---|
| prompt | Yes | Describe the output video. Refer to reference inputs as <Picture 1>..<Picture 9> and <Audio 1>..<Audio 3>. |
| video | Yes | URL of the source video. Up to 15 seconds are used; the control signal is extracted from it automatically, and the output keeps its duration and aspect ratio. |
| control_type | No | pose (default), depth, canny, soft_edge, lines, scribble or gray. |
| control_strength | No | How strongly the control constrains the output, 0 to 2. Default: 1. |
| reference_images | No | Optional reference images for the subject or style, up to 9. |
| reference_audios | No | Optional reference audio to guide audio generation, up to 3. |
| resolution | No | 480p, 540p, 768p or 1080p. Default: 480p. |
| generate_audio | No | Generate native audio (default), or false to keep the source video's audio track. |
| seed | No | Random seed. A negative value means a random seed will be used. |
| control_type | Keeps from the source | Good for |
|---|---|---|
| pose | Body, hand and face motion | Character animation, dance, performance transfer |
| depth | Spatial layout and camera movement | Changing the scene or style while keeping composition |
| canny | Precise edges | Faithful restyling of detailed shots |
| soft_edge | Softer outlines | Stylized re-rendering with more freedom |
| lines | Straight lines | Architecture, interiors, products |
| scribble | Loose sketch lines | Creative reinterpretation with rough structure |
| gray | Brightness structure | Colorizing, recoloring and relighting |
Cost = per-second rate × source video seconds + $0.02 per reference image or reference audio.
Source video seconds are rounded up, with a minimum of 3 and a maximum of 15 (only the first 15 seconds are used). The output has the same duration as the source.
| Resolution | Per second | 3s video | 5s video | 10s video | 15s video |
|---|---|---|---|---|---|
| 480p | $0.10 | $0.30 | $0.50 | $1.00 | $1.50 |
| 540p | $0.15 | $0.45 | $0.75 | $1.50 | $2.25 |
| 768p | $0.20 | $0.60 | $1.00 | $2.00 | $3.00 |
| 1080p | $0.40 | $1.20 | $2.00 | $4.00 | $6.00 |
Add-ons: reference image $0.02 each, reference audio $0.02 each. control_type, control_strength, generate_audio, prompt, and seed do not add separate charges.
| Configuration | Calculation | Cost |
|---|---|---|
| 5s video, 480p | 5 × $0.10 | $0.50 |
| 8s video, 768p + 1 reference image | 8 × $0.20 + $0.02 | $1.62 |
| 20s video (first 15s used), 1080p | 15 × $0.40 | $6.00 |
pose with a clear full-body reference image in reference_images and describe the character in the prompt.control_strength when the prompt should change more; raise it when the output drifts from the source.generate_audio to false to keep the original soundtrack, for example for dance videos.prompt and video are required.15 seconds of the source video are used; the output keeps the source duration and aspect ratio.reference_images supports up to 9 images and reference_audios up to 3.Grab a WaveSpeedAI API key, then call POST https://api.wavespeed.ai/api/v3/wavespeed-ai/minimax-h3/controlnet-union with your input as JSON. The endpoint returns a prediction id. Start polling the result endpoint around every 2 seconds, increase the interval for long-running tasks, and stop on any terminal status. On completed, read output values from data.outputs. Examples for Minimax H3 Controlnet Union below.
set -euo pipefail
: "${WAVESPEED_API_KEY:?Set WAVESPEED_API_KEY}"
REQUEST_BODY=$(cat <<'JSON'
{
"prompt": "A cinematic shot of a city at sunset, soft golden light",
"video": "https://interactive-examples.mdn.mozilla.net/media/cc0-videos/flower.mp4",
"control_type": "pose",
"control_strength": 1,
"resolution": "480p",
"generate_audio": true
}
JSON
)
# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
-X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/minimax-h3/controlnet-union" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $WAVESPEED_API_KEY" \
-d "$REQUEST_BODY")
TASK=$(printf '%s' "$SUBMIT_RESPONSE" | jq 'if has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "$TASK" | jq -r '.id')
if [ -z "$PREDICTION_ID" ] || [ "$PREDICTION_ID" = "null" ]; then
printf 'Submission response did not contain a prediction id
' >&2
exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/$PREDICTION_ID/result"
# 2. Poll until the prediction finishes.
while true; do
RESPONSE=$(curl --silent --show-error --fail-with-body "$RESULT_URL" \
-H "Authorization: Bearer $WAVESPEED_API_KEY")
RESULT=$(printf '%s' "$RESPONSE" | jq 'if has("data") then .data else . end')
STATUS=$(printf '%s' "$RESULT" | jq -r '.status')
case "$STATUS" in
completed) printf '%s\n' "$RESULT" | jq '.outputs'; break ;;
failed|cancelled|timeout|deleted) printf '%s\n' "$RESULT" | jq . >&2; exit 1 ;;
*) sleep 2 ;;
esac
doneconst submitUrl = "https://api.wavespeed.ai/api/v3/wavespeed-ai/minimax-h3/controlnet-union";
const apiKey = process.env.WAVESPEED_API_KEY;
if (!apiKey) throw new Error('Set WAVESPEED_API_KEY');
async function requestJson(url, options = {}) {
const response = await fetch(url, options);
if (!response.ok) throw new Error(await response.text());
return response.json();
}
// 1. Submit the prediction.
const body = await requestJson(submitUrl, {
method: "POST",
headers: {
"Authorization": `Bearer ${apiKey}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"prompt": "A cinematic shot of a city at sunset, soft golden light",
"video": "https://interactive-examples.mdn.mozilla.net/media/cc0-videos/flower.mp4",
"control_type": "pose",
"control_strength": 1,
"resolution": "480p",
"generate_audio": true
}),
});
const task = body.data ?? body;
if (!task.id) throw new Error("Submission response did not contain a prediction id");
const resultUrl = `https://api.wavespeed.ai/api/v3/predictions/${task.id}/result`;
// 2. Poll until the prediction finishes.
while (true) {
const resultBody = await requestJson(resultUrl, {
headers: { "Authorization": `Bearer ${apiKey}` },
});
const result = resultBody.data ?? resultBody;
if (result.status === "completed") {
console.log(result.outputs);
break;
}
if (["failed", "cancelled", "timeout", "deleted"].includes(result.status)) throw new Error(JSON.stringify(result));
await new Promise(resolve => setTimeout(resolve, 2000));
}import json
import os
import time
from urllib.request import Request, urlopen
api_key = os.environ["WAVESPEED_API_KEY"]
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
payload = {
"prompt": "A cinematic shot of a city at sunset, soft golden light",
"video": "https://interactive-examples.mdn.mozilla.net/media/cc0-videos/flower.mp4",
"control_type": "pose",
"control_strength": 1,
"resolution": "480p",
"generate_audio": True
}
def request_json(url, data=None):
request = Request(url, data=data, headers=headers, method="POST" if data else "GET")
with urlopen(request) as response:
return json.load(response)
# 1. Submit the prediction.
body = request_json("https://api.wavespeed.ai/api/v3/wavespeed-ai/minimax-h3/controlnet-union", json.dumps(payload).encode())
task = body.get("data", body)
if not task.get("id"):
raise RuntimeError("Submission response did not contain a prediction id")
result_url = f"https://api.wavespeed.ai/api/v3/predictions/{task['id']}/result"
# 2. Poll until the prediction finishes.
while True:
result_body = request_json(result_url)
result = result_body.get("data", result_body)
status = result.get("status")
if status == "completed":
print(result.get("outputs", []))
break
if status in {"failed", "cancelled", "timeout", "deleted"}:
raise RuntimeError(result)
time.sleep(2)Minimax H3 Controlnet Union is a WaveSpeedAI model for video editing, exposed as a REST API on WaveSpeedAI. MiniMax H3 Open Weights ControlNet Union generates a new video that follows the motion and composition of a source video. Pose, depth, edges, lines, scribble or grayscale structure is extracted from the source automatically and guides the output, optionally with reference images for the subject or style, with native stereo audio generated in the same pass. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing. You can call it programmatically or try it from the playground above.
POST your input parameters to the model's REST endpoint (shown in the API tab of this playground) with your WaveSpeedAI API key in the Authorization header. Submission returns a prediction ID. Poll the result endpoint starting around every 2 seconds, increase the interval for long-running tasks, and stop on any terminal status. The playground generates Python, JavaScript, and cURL examples for submitting requests and polling results. Full request/response shape is documented at https://wavespeed.ai/docs/docs-api/wavespeed-ai/minimax-h3-controlnet-union.
Minimax H3 Controlnet Union starts at $0.15 per run. That figure is the base price — the final charge scales with the parameters you set in the form (output size, length, count, references, or whatever knobs this model exposes), so a higher-quality or larger output costs more than a minimal one. The exact cost for your current input is shown live next to the Generate button before you submit, and the actual per-call charge is recorded on the prediction afterwards.
Key inputs: `prompt`, `video`, `resolution`, `seed`, `reference_images`, `control_strength`. The full JSON schema (types, defaults, allowed values) is rendered above the Generate button and mirrored in the API reference at https://wavespeed.ai/docs/docs-api/wavespeed-ai/minimax-h3-controlnet-union.
Sign up for a free WaveSpeedAI account to claim starter credits, copy your API key from /accesskey, then call the endpoint shown in the API tab of the playground. The playground also auto-generates a code sample in Python, JavaScript, or cURL for the parameters you've set.
Commercial usage rights depend on the model's license, set by its provider (WaveSpeedAI). Check the provider's applicable terms and WaveSpeedAI's Terms of Service before commercial use.