Hailuo Series

POST

tasks

Hailuo Series

curl --request POST \
  --url https://api.qingbo.ai/v1/tasks \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "<string>",
  "action": "<string>",
  "prompt": "<string>",
  "duration": 123,
  "resolution": "<string>",
  "first_frame_image": "<string>",
  "last_frame_image": "<string>",
  "prompt_optimizer": true,
  "fast_pretreatment": true,
  "watermark": true
}
'

import requests

url = "https://api.qingbo.ai/v1/tasks"

payload = {
    "model": "<string>",
    "action": "<string>",
    "prompt": "<string>",
    "duration": 123,
    "resolution": "<string>",
    "first_frame_image": "<string>",
    "last_frame_image": "<string>",
    "prompt_optimizer": True,
    "fast_pretreatment": True,
    "watermark": True
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: '<string>',
    action: '<string>',
    prompt: '<string>',
    duration: 123,
    resolution: '<string>',
    first_frame_image: '<string>',
    last_frame_image: '<string>',
    prompt_optimizer: true,
    fast_pretreatment: true,
    watermark: true
  })
};

fetch('https://api.qingbo.ai/v1/tasks', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.qingbo.ai/v1/tasks"

	payload := strings.NewReader("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"duration\": 123,\n  \"resolution\": \"<string>\",\n  \"first_frame_image\": \"<string>\",\n  \"last_frame_image\": \"<string>\",\n  \"prompt_optimizer\": true,\n  \"fast_pretreatment\": true,\n  \"watermark\": true\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.qingbo.ai/v1/tasks")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"duration\": 123,\n  \"resolution\": \"<string>\",\n  \"first_frame_image\": \"<string>\",\n  \"last_frame_image\": \"<string>\",\n  \"prompt_optimizer\": true,\n  \"fast_pretreatment\": true,\n  \"watermark\": true\n}")
  .asString();

MiniMax’s Hailuo video line. This page covers minimax-hailuo-02, minimax-hailuo-2.3 and minimax-hailuo-2.3-fast: all three go through the same task endpoint, all three have generate as their only action, and each turns text or a first-frame image into one clip of a fixed length — switching tiers means changing model and nothing else. For reference video, reference audio or a native soundtrack, use MiniMax H3.

Quick start

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: hailuo-demo-001" \
  -d '{
    "model": "minimax-hailuo-2.3",
    "action": "generate",
    "prompt": "Mist clears over a mountain temple at dawn, camera slowly pushing in",
    "duration": 6,
    "resolution": "1080p"
  }'

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-hailuo-02",
    "action": "generate",
    "prompt": "The subject slowly looks up, hair moving in the wind",
    "first_frame_image": "https://cdn.example.com/portrait.jpg",
    "duration": 10,
    "resolution": "768p"
  }'

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-hailuo-02",
    "action": "generate",
    "prompt": "Day turns into night, street lights coming on one by one",
    "first_frame_image": "https://cdn.example.com/day.jpg",
    "last_frame_image": "https://cdn.example.com/night.jpg",
    "duration": 10,
    "resolution": "768p"
  }'

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-hailuo-2.3-fast",
    "action": "generate",
    "prompt": "City at night, neon signs along the street, camera panning right",
    "first_frame_image": "https://cdn.example.com/city-night.jpg",
    "duration": 6,
    "resolution": "768p",
    "prompt_optimizer": true
  }'

A successful submission returns a task_id. Retrieve the result with GET /v1/tasks/{task_id}, wait inline with Prefer: wait, or configure a webhook.

Request parameters

string

required

One of minimax-hailuo-02, minimax-hailuo-2.3 or minimax-hailuo-2.3-fast.

string

default:"generate"

generate is the only action of all three models. The mode is not selected by action: prompt only is text-to-video, adding first_frame_image is first-frame video, and a first frame plus last_frame_image is first-and-last-frame video.

string

required

Description of the video: scene, subject, motion, style. minimax-hailuo-2.3 and minimax-hailuo-2.3-fast also accept camera-movement tags inside the prompt, see Examples.

integer

Clip length in seconds.

minimax-hailuo-02: 5 or 10, default 5
minimax-hailuo-2.3 and minimax-hailuo-2.3-fast: 6 or 10, default 6

string

Output resolution.

minimax-hailuo-02: 512p, 768p or 1080p, default 768p
minimax-hailuo-2.3 and minimax-hailuo-2.3-fast: 768p or 1080p, default 768p

string

URL of the first frame, used as the opening image of the clip. Required by minimax-hailuo-2.3-fast.

string

URL of the last frame; combine with first_frame_image for an A-to-B transition. Supported by minimax-hailuo-02 only.

boolean

default:"true"

Auto-optimize the prompt.

boolean

default:"false"

Shorten prompt-optimization time; faster, with slightly weaker optimization.

boolean

default:"false"

Add a watermark to the output.

The shared callback_url, callback_events, Prefer: wait, Idempotency-Key and cost-cap headers are documented in Submit Task. None of the three models accepts image_urls, video_urls, audio_urls, aspect_ratio, n, seed or negative_prompt; minimax-hailuo-2.3 / minimax-hailuo-2.3-fast do not accept last_frame_image. Anything beyond the parameters above returns 400 and is not billed.

Limits

Model	Condition	Limit
`minimax-hailuo-02`	Per request	Each task returns one video.
`minimax-hailuo-02`	`resolution: "512p"`	512p is only available with a first frame image (image-to-video).
`minimax-hailuo-02`	`resolution: "1080p"`	1080p supports 5 seconds only.
`minimax-hailuo-02`	`last_frame_image` sent	A last frame requires a first frame.
`minimax-hailuo-2.3` `minimax-hailuo-2.3-fast`	Per request	This model produces one result per task; n is not supported
`minimax-hailuo-2.3` `minimax-hailuo-2.3-fast`	`resolution: "1080p"`	1080p only supports 6-second duration
`minimax-hailuo-2.3-fast`	Per request	`first_frame_image` is required

minimax-hailuo-02 takes at most 2 images per request: one first frame and one last frame.

Pricing

Billed at a fixed price per clip for each resolution × duration tier, not per second: every allowed resolution and duration combination has its own price, so the cost of a call is known before you submit.

Model ID	Billing tiers
`minimax-hailuo-02`	`512p` 5 s / 10 s, `768p` 5 s / 10 s, `1080p` 5 s
`minimax-hailuo-2.3`	`768p` 6 s / 10 s, `1080p` 6 s
`minimax-hailuo-2.3-fast`	`768p` 6 s / 10 s, `1080p` 6 s

Without resolution or duration, the tier of the defaults (768p, 5 or 6 seconds) applies. prompt_optimizer, fast_pretreatment, watermark and the first / last frame images do not affect the price.

Per-model rates are in price_config on GET /v1/models, and in the console’s Model Market.What a single call actually cost is the cost field on the task response — an integer quota at 500,000 quota = 1 USD.

You are only billed for a video that is produced. Failed and cancelled tasks, and tasks that return no usable video, are refunded in full. The final charge is the cost field on the task response — an integer quota, not dollars.

Response

Task completed

{
  "task_id": "task-wave1788773955b950509106",
  "model": "minimax-hailuo-02",
  "action": "generate",
  "status": "completed",
  "progress": "100%",
  "created_at": 1788773955,
  "completed_at": 1788774046,
  "result": {
    "videos": [
      {
        "url": ["https://cdn.example.com/result.mp4"],
        "expires_at": 1789089508
      }
    ]
  },
  "billing_status": "settled",
  "cost": 58500,
  "urls": {
    "get": "https://api.qingbo.ai/v1/tasks/task-wave1788773955b950509106",
    "cancel": "https://api.qingbo.ai/v1/tasks/task-wave1788773955b950509106/cancel"
  }
}

result.videos holds the generated video and expires_at is when the link stops working (Unix seconds) — copy the file to your own storage before then. Full field reference in Query Task Status.

Available models

Model ID	Resolutions	Duration (s)	Modes	Notes
`minimax-hailuo-02`	`512p` `768p` (default) `1080p`	5 (default) / 10	Text, first frame, first and last frame	`512p` needs a first frame; `1080p` is 5 s only
`minimax-hailuo-2.3`	`768p` (default) `1080p`	6 (default) / 10	Text, first frame	Camera-movement tags
`minimax-hailuo-2.3-fast`	`768p` (default) `1080p`	6 (default) / 10	First frame (required)	Faster tier of 2.3, lower rate

Examples

minimax-hailuo-2.3 and minimax-hailuo-2.3-fast take camera-movement tags written as [tag] inside prompt. The tag labels are Chinese and must be written exactly as below; they can be mixed into the description and several can be used in one prompt:

Category	Tags
Pan	`[左移]` (left) `[右移]` (right)
Rise / descend	`[上升]` (rise) `[下降]` (descend)
Push / pull	`[推进]` (push in) `[拉远]` (pull out)
Rotate	`[左摇]` `[右摇]` `[上摇]` `[下摇]` (left, right, up, down)
Zoom	`[变焦推近]` (zoom in) `[变焦拉远]` (zoom out)
Other	`[晃动]` (shake) `[跟随]` (follow) `[固定]` (static)

Request body with camera tags

{
  "model": "minimax-hailuo-2.3",
  "action": "generate",
  "prompt": "[推进] A cobbled street in an old town, [变焦推近] rain dripping from the eaves",
  "duration": 6,
  "resolution": "1080p"
}

Hailuo Series

curl --request POST \
  --url https://api.qingbo.ai/v1/tasks \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "<string>",
  "action": "<string>",
  "prompt": "<string>",
  "duration": 123,
  "resolution": "<string>",
  "first_frame_image": "<string>",
  "last_frame_image": "<string>",
  "prompt_optimizer": true,
  "fast_pretreatment": true,
  "watermark": true
}
'

import requests

url = "https://api.qingbo.ai/v1/tasks"

payload = {
    "model": "<string>",
    "action": "<string>",
    "prompt": "<string>",
    "duration": 123,
    "resolution": "<string>",
    "first_frame_image": "<string>",
    "last_frame_image": "<string>",
    "prompt_optimizer": True,
    "fast_pretreatment": True,
    "watermark": True
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: '<string>',
    action: '<string>',
    prompt: '<string>',
    duration: 123,
    resolution: '<string>',
    first_frame_image: '<string>',
    last_frame_image: '<string>',
    prompt_optimizer: true,
    fast_pretreatment: true,
    watermark: true
  })
};

fetch('https://api.qingbo.ai/v1/tasks', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.qingbo.ai/v1/tasks"

	payload := strings.NewReader("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"duration\": 123,\n  \"resolution\": \"<string>\",\n  \"first_frame_image\": \"<string>\",\n  \"last_frame_image\": \"<string>\",\n  \"prompt_optimizer\": true,\n  \"fast_pretreatment\": true,\n  \"watermark\": true\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.qingbo.ai/v1/tasks")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"duration\": 123,\n  \"resolution\": \"<string>\",\n  \"first_frame_image\": \"<string>\",\n  \"last_frame_image\": \"<string>\",\n  \"prompt_optimizer\": true,\n  \"fast_pretreatment\": true,\n  \"watermark\": true\n}")
  .asString();

Text Generation

Image Generation

Video Generation

Audio

Tools & Embeddings

Task Management

Account & Usage

Quick start

Request parameters

Limits

Pricing

Response

Available models

Examples

​Quick start

​Request parameters

​Limits

​Pricing

​Response

​Available models

​Examples

​Related

Quick start

Request parameters

Limits

Pricing

Response

Available models

Examples

Related