tuning (training)-Video generation model fine-tuning API reference

Fine-tune Wan image-to-video models through the Model Studio API. This reference covers each API operation in the fine-tuning workflow -- from dataset upload through model deployment.

Workflow overview

A typical fine-tuning workflow follows this sequence:

Upload a dataset -- Upload training data (.zip) and get a file ID.
Create a fine-tuning job -- Start training with a base model and the uploaded dataset.
Query job status -- Poll until training completes (several hours).
Deploy the model -- Publish the fine-tuned model as an API service.
Query deployment status -- Poll until deployment is ready (5--10 minutes).
Generate videos -- Call the deployed model.

Optional operations:

Manage checkpoints -- List, preview, and export specific training checkpoints.
Manage jobs -- List, cancel, or delete fine-tuning jobs.
Delete deployments -- Take a deployed model offline.

Before you begin

Region: Only the Singapore region is supported under the international deployment mode. Use an API key from this region.

API key: Create an API key and set it as an environment variable.

Background reading: Read the fine-tuning overview to understand supported models, data format, fine-tuning steps, and billing.

Sample datasets:

Mode	Training set	Validation set
Image-to-video (first frame)	Download	Download
Image-to-video (first and last frames)	Download	Download

Supported base models

Mode	Model ID
Image-to-video (first frame)	`wan2.6-i2v`, `wan2.5-i2v-preview`, `wan2.2-i2v-flash`
Image-to-video (first and last frames)	`wan2.2-kf2v-flash`

Upload a dataset

Note

When using the API, the zip package size must be ≤ 1GB.

Upload a .zip dataset and get a file ID to create fine-tuning jobs.

Request

POST https://dashscope-intl.aliyuncs.com/compatible-mode/v1/files
Content-Type: multipart/form-data

Parameter	Location	Type	Required	Description	Example
`file`	Body (form-data)	file	Yes	Local `.zip` dataset. Pass as `files=@" "`.	`@"./wan-i2v-training-dataset.zip"`
`purpose`	Body (form-data)	string	Yes	Fixed to `fine-tune`.	`fine-tune`

Response

Field	Type	Description	Example
`id`	string	File ID for creating fine-tuning jobs.	`file-ft-b2416bacc4d742xxxx`
`object`	string	Upload type.	`file`
`bytes`	integer	File size in bytes.	`73310369`
`filename`	string	File name.	`wan-i2v-training-dataset.zip`
`purpose`	string	Fixed to `fine-tune`.	`fine-tune`
`status`	string	Upload status. `processed` means the upload succeeded.	`processed`
`created_at`	integer	Creation time as a UNIX timestamp.	`1766127125`

Example

Request

HELPCODEESCAPE-bash
curl --location --request POST 'https://dashscope-intl.aliyuncs.com/compatible-mode/v1/files' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--form 'file=@"./wan-i2v-training-dataset.zip"' \
--form 'purpose="fine-tune"'

Response

Copy the id value for the next step.

HELPCODEESCAPE-json
{
    "id": "file-ft-b2416bacc4d742xxxx",
    "object": "file",
    "bytes": 73310369,
    "filename": "wan-i2v-training-dataset.zip",
    "purpose": "fine-tune",
    "status": "processed",
    "created_at": 1766127125
}

Next step: Use the file ID to create a fine-tuning job.

Create a fine-tuning job

Start training with a base model and uploaded dataset.

Request

POST https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes
Content-Type: application/json

Parameter	Location	Type	Required	Description	Example
`model`	Body	string	Yes	Base model ID (see Supported base models ).	`wan2.5-i2v-preview`
`training_file_ids`	Body	array[string]	Yes	Training set file IDs. Accepts multiple IDs.	`\["file-ft-b2416bacc4d742xxxx"\]`
`validation_file_ids`	Body	array[string]	No	Validation set file IDs (if omitted, training set is auto-split).	-
`training_type`	Body	string	Yes	Fine-tuning method. Only `efficient_sft` (LoRA) is supported.	`efficient_sft`
`hyper_parameters`	Body	object	No	Hyperparameter settings. See Hyperparameters .	See below

Hyperparameters

Defaults work well for first runs. Adjust n_epochs or learning_rate if needed.

Quick-start defaults

Model	`batch_size`	`n_epochs`	`learning_rate`	`max_pixels`	`eval_epochs`
wan2.6-i2v	2	400	2e-5	36864	50
wan2.5-i2v-preview	2	400	2e-5	36864	50
wan2.2-i2v-flash	4	400	2e-5	262144	50
wan2.2-kf2v-flash	4	400	2e-5	262144	50

Full parameter reference

Parameter	Type	Required	Description	Default
`batch_size`	int	Yes	Samples per training step (use model default above).	2 or 4
`n_epochs`	int	Yes	Training epochs. Total steps = `n_epochs` × `ceil(dataset_size / batch_size)` (minimum 800 steps required). For 5 samples with `batch_size=2`: `ceil(5/2)=3` steps/epoch, so minimum `n_epochs = ceil(800/3) = 267`.	400
`learning_rate`	float	Yes	Weight update magnitude (too high degrades model; too low has minimal effect).	2e-5
`eval_epochs`	int	Yes	Epochs between evaluations (must be ≥ `n_epochs / 10`).	50
`max_pixels`	int	Yes	Maximum pixels (width × height). Videos exceeding this are downscaled. wan2.6/2.5: 16384--36864; wan2.2: 65536--262144.	36864 or 262144
`split`	float	No	Training-validation split ratio. Takes effect only when `validation_file_ids` is not provided. `0.9` means 90% training, 10% validation.	0.9
`max_split_val_dataset_sample`	int	No	Maximum validation samples when auto-splitting. `Validation size = min(total * (1 - split), max_split_val_dataset_sample)`. For example, 100 samples with `split=0.9` and this value set to 5: calculated validation = 10, capped to 5.	5
`save_total_limit`	int	No	Maximum number of checkpoints to keep. Only the last N checkpoints are retained.	20
`lora_rank`	int	No	LoRA low-rank matrix dimension. Larger values increase fitting capacity but slow training. Must be a power of 2 (16, 32, 64).	32
`lora_alpha`	int	No	LoRA weight scaling factor. Adjusts how much the fine-tuned parameters influence the base model weights. Must be a power of 2 (16, 32, 64).	32

Response

Field	Type	Description	Example
`request_id`	string	Request ID.	`0eb05b0c-02ba-414a-9d0c-xxxxxxxxx`
`output.job_id`	string	Fine-tuning job ID. Required to query job status.	`ft-202511111122-xxxx`
`output.job_name`	string	Job name.	`ft-202511111122-xxxx`
`output.status`	string	Job status. See Job status values .	`PENDING`
`output.finetuned_output`	string	Fine-tuned model name. Required for deployment.	`wan2.5-i2v-preview-ft-202511111122-xxxx`
`output.model`	string	Base model used.	`wan2.5-i2v-preview`
`output.base_model`	string	Base model used.	`wan2.5-i2v-preview`
`output.training_file_ids`	array	Training set file IDs.	`\["file-ft-b2416bacc4d742xxxx"\]`
`output.validation_file_ids`	array	Validation set file IDs. Empty if none were uploaded.	`\[\]`
`output.hyper_parameters`	object	Hyperparameters applied to this job.	`{...}`
`output.training_type`	string	Fixed to `efficient_sft`.	`efficient_sft`
`output.create_time`	string	Job creation time.	`2025-11-11 11:22:22`
`output.workspace_id`	string	Workspace ID. See Get a workspace ID.	`llm-xxxxxxxxx`
`output.user_identity`	string	Alibaba Cloud account ID.	`12xxxxxxx`
`output.modifier`	string	Account ID of the last modifier.	`12xxxxxxx`
`output.creator`	string	Account ID of the creator.	`12xxxxxxx`
`output.group`	string	Job group.	`llm`
`output.max_output_cnt`	integer	Maximum checkpoints to save. Equals `save_total_limit`.	`20`

Job status values

Status	Meaning
`PENDING`	Pending.
`QUEUING`	Queued (one job at a time).
`RUNNING`	In progress.
`CANCELING`	Canceling.
`SUCCEEDED`	Completed.
`FAILED`	Failed.
`CANCELED`	Canceled.

Examples

Replace <training-dataset-file-id> with the file ID from the Upload a dataset response.

wan2.6-i2v

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
    "model":"wan2.6-i2v",
    "training_file_ids":[
        "&lt;training-dataset-file-id&gt;"
    ],
    "training_type":"efficient_sft",
    "hyper_parameters":{
        "n_epochs":400,
        "batch_size":2,
        "learning_rate":2e-5,
        "split":0.9,
        "eval_epochs": 50,
        "max_pixels": 36864
    }
}'

wan2.5-i2v-preview

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
    "model":"wan2.5-i2v-preview",
    "training_file_ids":[
        "&lt;training-dataset-file-id&gt;"
    ],
    "training_type":"efficient_sft",
    "hyper_parameters":{
        "n_epochs":400,
        "batch_size":2,
        "learning_rate":2e-5,
        "split":0.9,
        "eval_epochs": 50,
        "max_pixels": 36864
    }
}'

wan2.2-i2v-flash

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
    "model":"wan2.2-i2v-flash",
    "training_file_ids":[
        "&lt;training-dataset-file-id&gt;"
    ],
    "training_type":"efficient_sft",
    "hyper_parameters":{
        "n_epochs":400,
        "batch_size":4,
        "learning_rate":2e-5,
        "split":0.9,
        "eval_epochs": 50,
        "max_pixels": 262144
    }
}'

wan2.2-kf2v-flash (first and last frames)

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
    "model":"wan2.2-kf2v-flash",
    "training_file_ids":[
        "&lt;training-dataset-file-id&gt;"
    ],
    "training_type":"efficient_sft",
    "hyper_parameters":{
        "n_epochs":400,
        "batch_size":4,
        "learning_rate":2e-5,
        "split":0.9,
        "eval_epochs": 50,
        "max_pixels": 262144
    }
}'

Multiple training and validation sets

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
    "model":"wan2.5-i2v-preview",
    "training_file_ids":[
        "&lt;training-set-file-id-1&gt;",
        "&lt;training-set-file-id-2&gt;"
    ],
    "validation_file_ids":[
         "&lt;validation-set-file-id-1&gt;",
         "&lt;validation-set-file-id-2&gt;"
    ],
    "training_type":"efficient_sft",
    "hyper_parameters":{
        "n_epochs":400,
        "batch_size":2,
        "learning_rate":2e-5,
        "split":0.9,
        "eval_epochs": 50,
        "max_pixels": 36864
    }
}'

Response

Copy output.job_id (to query status) and output.finetuned_output (to deploy the model).

HELPCODEESCAPE-json
{
    "request_id": "0eb05b0c-02ba-414a-9d0c-xxxxxxxxx",
    "output": {
        "job_id": "ft-202511111122-xxxx",
        "job_name": "ft-202511111122-xxxx",
        "status": "PENDING",
        "finetuned_output": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "model": "wan2.5-i2v-preview",
        "base_model": "wan2.5-i2v-preview",
        "training_file_ids": [
            "xxxxxxxxxxxx"
        ],
        "validation_file_ids": [],
        "hyper_parameters": {
            "n_epochs": 400,
            "batch_size": 2,
            "learning_rate": 2.0E-5,
            "split": 0.9,
            "eval_epochs": 50
        },
        "training_type": "efficient_sft",
        "create_time": "2025-11-11 11:22:22",
        "workspace_id": "llm-xxxxxxxxx",
        "user_identity": "12xxxxxxx",
        "modifier": "12xxxxxxx",
        "creator": "12xxxxxxx",
        "group": "llm",
        "max_output_cnt": 20
    }
}

Next step: Query the job status and wait for training to complete.

Query job status

Check job progress. Poll until status is SUCCEEDED.

Training takes several hours (varies by model and dataset size).

Request

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}

Parameter	Location	Type	Required	Description	Example
`job_id`	Path	string	Yes	Fine-tuning job ID from the create job response.	`ft-202511111122-xxxx`

Response

All fields from the create job response are returned, plus:

Field	Type	Description	Example
`output.end_time`	string	Job completion time.	`2025-11-11 16:49:01`
`output.output_cnt`	integer	Actual number of saved checkpoints.	`8`
`output.usage`	integer	Total training tokens (for billing).	`432000`

Example

Replace <fine-tuning-job-id> with the job_id from the create job response.

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json'

Response

When status is SUCCEEDED, training is complete. The usage field shows total tokens consumed for billing.

HELPCODEESCAPE-json
{
    "request_id": "9bbb953c-bef2-4b59-9fc5-xxxxxxxxx",
    "output": {
        "job_id": "ft-202511111122-xxxx",
        "job_name": "ft-202511111122-xxxx",
        "status": "SUCCEEDED",
        "finetuned_output": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "model": "wan2.5-i2v-preview",
        "base_model": "wan2.5-i2v-preview",
        "training_file_ids": [
            "xxxxxxxxxxxx"
        ],
        "validation_file_ids": [],
        "hyper_parameters": {
            "n_epochs": 400,
            "batch_size": 2,
            "learning_rate": 2.0E-5,
            "split": 0.9,
            "eval_epochs": 50
        },
        "training_type": "efficient_sft",
        "create_time": "2025-11-11 11:22:22",
        "workspace_id": "llm-xxxxxxxxx",
        "user_identity": "xxxxxxxxx",
        "modifier": "xxxxxxxxx",
        "creator": "xxxxxxxxx",
        "end_time": "2025-11-11 16:49:01",
        "group": "llm",
        "usage": 432000,
        "max_output_cnt": 8,
        "output_cnt": 8
    }
}

Next step: Deploy the model or manage checkpoints to select a specific checkpoint.

Deploy a model

Publish a fine-tuned model as an API service after job status is SUCCEEDED.

Request

HELPCODEESCAPE-http
POST https://dashscope-intl.aliyuncs.com/api/v1/deployments
Content-Type: application/json

Parameter	Location	Type	Required	Description	Example
`model_name`	Body	string	Yes	Model to deploy (`output.finetuned_output` from job response or `output\[\].model_name` from checkpoint).	`wan2.5-i2v-preview-ft-202511111122-xxxx`
`capacity`	Body	integer	Yes	Model instances (set to `1`).	`1`
`plan`	Body	string	Yes	Fixed to `lora`.	`lora`
`aigc_config`	Body	object	Yes	Prompt configuration. See below.	-

Prompt configuration (aigc_config)

Parameter	Type	Required	Description
`use_input_prompt`	boolean	Yes	`false` (default): The system ignores the prompt in API requests and auto-generates one from the image and the templates below. `true`: The system uses the prompt from the API request directly, and ignores the template parameters.
`prompt`	string	Yes	Preset prompt template (active when `use_input_prompt=false`). System analyzes input image, fills subject/environment descriptions, and appends effect instructions. Overrides prompts from invocation API.
`lora_prompt_default`	string	Yes	Fallback template (active when `use_input_prompt=false` and `prompt` generation fails).

How to write the prompt template When use_input_prompt=false, the system auto-generates a complete prompt from the input image. Configure the prompt parameter so callers only need to upload an image.

Structure: Instruction + Output template + Example

Provide a video description based on the image content. The description must
include "Then the s86b5p money rain effect begins, and countless enormous US
dollar bills (beige with dark green patterns) pour down like a torrential rain,
densely hitting and surrounding her (him). The bills continue to fall as the
camera slowly zooms in. She (he) stretches her (his) arms upward, neck slightly
tilted back, with a surprised expression, completely immersed in this wild
money rain".

Output template:
The video begins by showing [subject description]. [Environment description].
Then the s86b5p money rain effect begins, and countless enormous US dollar bills
(beige with dark green patterns) pour down like a torrential rain, densely
hitting and surrounding her (him). The bills continue to fall as the camera
slowly zooms in. She (he) stretches her (his) arms upward, neck slightly tilted
back, with a surprised expression, completely immersed in this wild money rain.

Example:
The video begins by showing a young woman on a beach. Her hair is wet, dark
brown, curly, and slightly messy. She has a bright smile on her face. The
background shows rough waves and distant mountains. Then the s86b5p money rain
effect begins, and countless enormous US dollar bills (beige with dark green
patterns) pour down like a torrential rain, densely hitting and surrounding her.
The bills continue to fall as the camera slowly zooms in. She stretches her arms
upward, neck slightly tilted back, with a surprised expression, completely
immersed in this wild money rain.

For details on writing video prompts, see Video annotation: Write prompts for videos.

Response

Field	Type	Description	Example
`request_id`	string	Request ID.	`0eb05b0c-02ba-414a-9d0c-xxxxxxxxx`
`output.deployed_model`	string	Deployed model ID. Required to query deployment status and call the model.	`wan2.5-i2v-preview-ft-202511111122-xxxx`
`output.model_name`	string	Model ID name.	`wan2.5-i2v-preview-ft-202511111122-xxxx`
`output.status`	string	Deployment status. See Deployment status values .	`PENDING`
`output.base_model`	string	Base model used.	`wan2.5-i2v-preview`
`output.gmt_create`	string	Task creation time.	`2025-11-11T17:46:53.294`
`output.gmt_modified`	string	Task update time.	`2025-11-11T17:46:53.294`
`output.workspace_id`	string	Workspace ID. See Get a workspace ID.	`llm-xxxxxxxxx`
`output.charge_type`	string	Billing method. `post_paid` = pay-as-you-go.	`post_paid`
`output.creator`	string	Creator's account ID.	`12xxxxxxx`
`output.modifier`	string	Modifier's account ID.	`12xxxxxxx`
`output.plan`	string	Deployment method.	`lora`

Examples

Replace <model-name> with the finetuned_output value from the create job response (or the model_name from an exported checkpoint).

Image-to-video (first frame)

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/deployments' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
    "model_name": "&lt;model-name&gt;",
    "aigc_config": {
        "use_input_prompt": false,
        "prompt": "Provide a video description based on the image content. The description must include the phrase \"Then the s86b5p money rain effect begins, countless huge-sized US dollar bills (beige background/dark green patterns) pour down like a torrential rain, densely hitting and surrounding her. The bills continue to fall, she/he stretches her/his arms upward, neck slightly tilted back, expression surprised, completely immersed in this wild money rain.\"\nOutput Template:\nThe video begins with a shot of [subject description]. [Environment description]. Then the s86b5p money rain effect begins, countless huge-sized US dollar bills (beige background/dark green patterns) pour down like a torrential rain, densely hitting and surrounding her. The bills continue to fall, she/he stretches her/his arms upward, neck slightly tilted back, expression surprised, completely immersed in this wild money rain.\nExample:\nThe video begins showing a young woman standing in front of a brick wall covered with ivy. She has long, smooth reddish-brown hair, wearing a white sleeveless dress, a shiny silver necklace, and a smile on her face. The brick wall in the background is covered with green vines, appearing rustic and natural. Then the s86b5p money rain effect begins, countless huge-sized US dollar bills (beige background/dark green patterns) pour down like a torrential rain, densely hitting and surrounding her. The bills continue to fall, she stretches her arms upward, neck slightly tilted back, expression surprised, completely immersed in this wild money rain.",
        "lora_prompt_default": "Then the s86b5p money rain effect begins, countless huge-sized US dollar bills (beige background/dark green patterns) pour down like a torrential rain, densely hitting and surrounding her. The bills continue to fall, she/he stretches her/his arms upward, neck slightly tilted back, expression surprised, completely immersed in this wild money rain."
    },
    "capacity": 1,
    "plan": "lora"
}'

Image-to-video (first and last frames)

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/deployments' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
    "model_name": "&lt;model-name&gt;",
    "aigc_config": {
        "use_input_prompt": false,
        "prompt": "Provide a video description based on the image content. The description must include the phrase \"Then she/he begins the s86b5p transformation.\"\nOutput Template:\nThe video begins with a shot of [subject description]. [Environment description]. Then she/he begins the s86b5p transformation.\nExample:\nThe video begins with a young woman in an outdoor setting. She has short, curly dark brown hair and a friendly smile. She is wearing a black Polo shirt with colorful floral embroidery. The background features green vegetation and distant mountains. Then she begins the s86b5p transformation.",
        "lora_prompt_default": "Then she/he begins the s86b5p transformation."
    },
    "capacity": 1,
    "plan": "lora"
}'

Response

output.status of PENDING means deployment is in progress.

HELPCODEESCAPE-json
{
    "request_id": "96020b2e-9072-4c8a-9981-xxxxxxxxx",
    "output": {
        "deployed_model": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "gmt_create": "2025-11-11T17:46:53.294",
        "gmt_modified": "2025-11-11T17:46:53.294",
        "status": "PENDING",
        "model_name": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "base_model": "wan2.5-i2v-preview",
        "workspace_id": "llm-xxxxxxxxx",
        "charge_type": "post_paid",
        "creator": "12xxxxxxx",
        "modifier": "12xxxxxxx",
        "plan": "lora"
    }
}

Next step: Query deployment status and wait for status to become RUNNING.

Query deployment status

Check if a deployed model is ready. Poll until status is RUNNING.

Deployment takes 5--10 minutes.

Request

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/deployments/{deployed_model}

Parameter	Location	Type	Required	Description	Example
`deployed_model`	Path	string	Yes	Deployed model ID from the deploy model response.	`wan2.5-i2v-preview-ft-202511111122-xxxx`

Response

Same fields as the deploy model response, plus additional status values.

Deployment status values

Status	Meaning
`PENDING`	Deploying.
`RUNNING`	Ready to accept requests.
`ARREARS_DOWN`	Stopped due to overdue payment.
`ARREARS_RECOVERING`	Resuming after overdue payment.
`FAILED`	Deployment failed.
`OFFLINING`	Taking service offline.
`UPDATING`	Upgrading or downgrading.
`UPDATING_FAILED`	Upgrade or downgrade failed.

Example

Replace <deployed-model> with output.deployed_model from the deploy response.

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/deployments/&lt;deployed-model&gt;' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json'

Response

When status is RUNNING, the model is deployed and ready for requests.

HELPCODEESCAPE-json
{
    "request_id": "66d15f35-0772-409f-bc70-xxxxxxxxx",
    "output": {
        "deployed_model": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "gmt_create": "2025-11-11T17:46:53",
        "gmt_modified": "2025-11-11T18:02:24",
        "status": "RUNNING",
        "model_name": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "base_model": "wan2.5-i2v-preview",
        "workspace_id": "llm-xxxxxxxxx",
        "charge_type": "post_paid",
        "creator": "12xxxxxxx",
        "modifier": "12xxxxxxxx",
        "plan": "lora"
    }
}

Next step: Call the model to generate a video.

Checkpoint management

After training, list checkpoints, preview validation results, export checkpoints, and deploy them.

List checkpoints

Returns checkpoints with successful validation previews (failed checkpoints excluded).

Available only after training completes.

Request

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}/validation-results

Parameter	Location	Type	Required	Description	Example
`job_id`	Path	string	Yes	Fine-tuning job ID.	`ft-202511111122-xxxx`

Response

Field	Type	Description	Example
`request_id`	string	Request ID.	`0eb05b0c-02ba-414a-9d0c-xxxxxxxxx`
`output`	array[object]	List of checkpoints.	-
`output\[\].checkpoint`	string	Checkpoint name.	`checkpoint-160`

Example

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;/validation-results' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json'

Response

HELPCODEESCAPE-json
{
    "request_id": "da1310f5-5a21-4e29-99d4-xxxxxx",
    "output": [
        {
            "checkpoint": "checkpoint-160"
        },
        {
            "checkpoint": "checkpoint-20"
        },
        {
            "checkpoint": "checkpoint-40"
        },
        {
            "checkpoint": "checkpoint-60"
        }
    ]
}

Get validation details for a checkpoint

Return validation video for a specific checkpoint.

Request

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}/validation-details/{checkpoint}?page_no=1&page_size=10

Parameter	Location	Type	Required	Description	Example
`job_id`	Path	string	Yes	Fine-tuning job ID.	`ft-202511111122-xxxx`
`checkpoint`	Path	string	Yes	Checkpoint name.	`checkpoint-160`
`page_no`	Query	integer	No	Page number. Default: `1`.	`1`
`page_size`	Query	integer	No	Results per page. Default: `10`.	`10`

Response

Field	Type	Description	Example
`request_id`	string	Request ID.	`375b3ad0-d3fa-451f-b629-xxxxxxx`
`output.page_no`	integer	Page number.	`1`
`output.page_size`	integer	Results per page.	`10`
`output.total`	integer	Total validation entries.	`1`
`output.list\[\].video_path`	string	Video URL (valid 24 hours---download promptly).	`https://finetune-swap-wulanchabu.oss-cn-wulanchabu.aliyuncs.com/xxx.mp4?Expires=xxxx`
`output.list\[\].prompt`	string	Prompt from the `data.jsonl` annotation file.	`The video begins with a young man...`
`output.list\[\].first_frame_path`	string	Validation image URL.	`https://finetune-swap-wulanchabu.oss-cn-wulanchabu.aliyuncs.com/xxx.jpeg`

Example

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;/validation-details/&lt;checkpoint&gt;?page_no=1&page_size=10' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY"

Response

Download within 24 hours before video_path expires.

HELPCODEESCAPE-json
{
    "request_id": "375b3ad0-d3fa-451f-b629-xxxxxxx",
    "output": {
        "page_no": 1,
        "page_size": 10,
        "total": 1,
        "list": [
            {
                "video_path": "https://finetune-swap-wulanchabu.oss-cn-wulanchabu.aliyuncs.com/xxx.mp4?Expires=xxxx",
                "prompt": "The video begins with a young man sitting in a cafe. He is wearing a beige Polo shirt, looking focused and slightly contemplative, with his fingers gently touching his chin. In front of him is a cup of hot coffee. The background is a wall with wooden stripes and a decorative sign. Then the s86b5p money rain effect begins, and countless enormous US dollar bills (beige with dark green patterns) pour down like a torrential rain, densely hitting and surrounding him. The bills continue to fall as he stretches his arms upward, neck slightly tilted back, with a surprised expression, completely immersed in this wild money rain.",
                "first_frame_path": "https://finetune-swap-wulanchabu.oss-cn-wulanchabu.aliyuncs.com/xxx.jpeg"
            }
        ]
    }
}

Export a checkpoint

Export a checkpoint as a deployable model.

Request

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}/export/{checkpoint}?model_name={model_name}

Parameter	Location	Type	Required	Description	Example
`job_id`	Path	string	Yes	Fine-tuning job ID.	`ft-202511111122-xxxx`
`checkpoint`	Path	string	Yes	Checkpoint name.	`checkpoint-160`
`model_name`	Query	string	Yes	Console display name (must be globally unique). Use Chinese characters, letters, digits, underscores, or hyphens. For deployment, use `output\[\].model_name` from the exported model details response.	`my-checkpoint-160`

Response

Field	Type	Description	Example
`request_id`	string	Request ID.	`0eb05b0c-02ba-414a-9d0c-xxxxxxxxx`
`output`	boolean	`true`: request submitted; `false`: failed (retry).	`true`

Example

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;/export/&lt;checkpoint&gt;?model_name=&lt;console-display-name&gt;' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY"

Response

HELPCODEESCAPE-json
{
    "request_id": "0817d1ed-b6b6-4383-9650-xxxxx",
    "output": true
}

Next step: Get exported model details to retrieve the model_name for deployment.

Get exported model details

Return export status and model_name for deployment.

Request

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}/checkpoints

Parameter	Location	Type	Required	Description	Example
`job_id`	Path	string	Yes	Fine-tuning job ID.	`ft-202511111122-xxxx`

Response

Field	Type	Description	Example
`request_id`	string	Request ID.	`0eb05b0c-02ba-414a-9d0c-xxxxxxxxx`
`output`	array[object]	Checkpoint details.	-
`output\[\].create_time`	string	Creation time.	`2025-11-11T13:27:29`
`output\[\].job_id`	string	Fine-tuning job ID.	`ft-202511111122-xxxx`
`output\[\].checkpoint`	string	Checkpoint name.	`checkpoint-160`
`output\[\].full_name`	string	Full checkpoint ID.	`ft-202511111122-496e-checkpoint-160`
`output\[\].model_name`	string	Model name for deployment (returned when `status` is `SUCCEEDED`).	`wan2.5-i2v-preview-ft-202511111122-xxxx-c160`
`output\[\].model_display_name`	string	Display name.	`wan2.5-i2v-preview-ft-202511111122-xxxx`
`output\[\].status`	string	Export status. See below.	`SUCCEEDED`

Export status values

Status	Meaning
`PENDING`	Queued for export.
`PROCESSING`	Exporting.
`SUCCEEDED`	Export completed.
`FAILED`	Export failed.
`UNSUPPORTED`	Export not supported.

Example

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;/checkpoints' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY"

Response

When target checkpoint status is SUCCEEDED, save its model_name for deployment.

HELPCODEESCAPE-json
{
    "request_id": "b0e33c6e-404b-4524-87ac-xxxxxx",
    "output": [
        {
            "create_time": "2025-11-11T13:42:31",
            "full_name": "ft-202511111122-496e-checkpoint-180",
            "job_id": "ft-202511111122-496e",
            "checkpoint": "checkpoint-180",
            "status": "PENDING"
        },
        {
            "create_time": "2025-11-11T13:27:29",
            "full_name": "ft-202511111122-496e-checkpoint-160",
            "job_id": "ft-202511111122-496e",
            "checkpoint": "checkpoint-160",
            "model_name": "wan2.5-i2v-preview-ft-202511111122-xxxx-c160",
            "model_display_name": "wan2.5-i2v-preview-ft-202511111122-xxxx",
            "status": "SUCCEEDED"
        }
    ]
}

Deploy and call an exported checkpoint

After exporting a checkpoint and getting its model_name:

Deploy the model -- pass the exported model_name as the model_name parameter.
Query deployment status -- wait for RUNNING.
Call the model to generate a video.

Fine-tuning job management

Query job logs

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}/logs

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;/logs' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY"

Response

HELPCODEESCAPE-json
{
    "request_id": "b7ecb456-6dd1-4f35-a581-xxxxxx",
    "output": {
        "total": 25,
        "logs": [
            "2025-11-11 11:23:37,315 - INFO - data process succeeded, start to fine-tune",
            " Actual number of consumed tokens is 215040 !",
            " Actual number of consumed tokens is 419840 !",
            " Actual number of consumed tokens is 624640 !",
            "2025-11-11 16:31:40,760 - INFO - fine-tuned output got, start to transfer it for inference",
            "2025-11-11 16:32:29,162 - INFO - transfer for inference succeeded, start to deliver it for inference",
            "2025-11-11 16:40:28,784 - INFO - start to save checkpoint",
            "2025-11-11 16:49:01,738 - INFO - finetune-job succeeded",
            "2025-11-11 16:49:02,234 - INFO - ##FT_COMPLETE##"
        ]
    }
}

List fine-tuning jobs

HELPCODEESCAPE-http
GET https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes

Request

HELPCODEESCAPE-bash
curl --location 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY"

Response

HELPCODEESCAPE-json
{
    "request_id": "bf4d3475-f50c-42e2-a263-xxxxxxxxx",
    "output": {
        "page_no": 1,
        "page_size": 10,
        "total": 1,
        "jobs": [
            {
                "job_id": "ft-202511111122-xxxx",
                "job_name": "ft-202511111122-xxxx",
                "status": "SUCCEEDED",
                "finetuned_output": "wan2.5-i2v-preview-ft-202511111122-xxxx",
                "model": "wan2.5-i2v-preview",
                "base_model": "wan2.5-i2v-preview",
                "training_file_ids": [
                    "xxxxxxxxx"
                ],
                "validation_file_ids": [],
                "hyper_parameters": {
                    "n_epochs": 400,
                    "batch_size": 2,
                    "learning_rate": 2.0E-5,
                    "split": 0.9,
                    "eval_epochs": 50
                },
                "training_type": "efficient_sft",
                "create_time": "2025-11-11 11:22:22",
                "workspace_id": "llm-xxxxxxxxx",
                "user_identity": "xxxxxxxxx",
                "modifier": "xxxxxxxxx",
                "creator": "xxxxxxxxx",
                "end_time": "2025-11-11 16:49:01",
                "group": "llm",
                "usage": 432000,
                "max_output_cnt": 8,
                "output_cnt": 8
            }
        ]
    }
}

Cancel a fine-tuning job

POST https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}/cancel

Request

HELPCODEESCAPE-bash
curl --location --request POST 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;/cancel' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json'

Response

HELPCODEESCAPE-json
{
    "request_id": "d8dab938-e32e-40bf-83ab-xxxxxx",
    "output": {
        "status": "success"
    }
}

Delete a fine-tuning job

DELETE https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/{job_id}

Request

HELPCODEESCAPE-bash
curl --location --request DELETE 'https://dashscope-intl.aliyuncs.com/api/v1/fine-tunes/&lt;fine-tuning-job-id&gt;' \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header 'Content-Type: application/json'

Response

HELPCODEESCAPE-json
{
    "request_id": "1301136c-12f2-4504-880a-xxxxxx",
    "output": {
        "status": "success"
    }
}

Deployed model management

Delete a deployed model

HELPCODEESCAPE-http
DELETE https://dashscope-intl.aliyuncs.com/api/v1/deployments/{deployed_model}

Important

Immediately takes model offline (cannot be undone). Model becomes unavailable and billing stops.

Request

HELPCODEESCAPE-bash
curl --request DELETE 'https://dashscope-intl.aliyuncs.com/api/v1/deployments/&lt;deployed-model&gt;' \
    --header "Authorization: Bearer $DASHSCOPE_API_KEY" \
    --header 'Content-Type: application/json'

Response

status: DELETING confirms deletion accepted.

HELPCODEESCAPE-json
{
    "request_id": "c2ed2aa2-39b8-4a86-b79e-xxxxxx",
    "output": {
        "deployed_model": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "gmt_create": "2025-11-11T17:46:53",
        "gmt_modified": "2025-12-22T11:18:27.532",
        "status": "DELETING",
        "model_name": "wan2.5-i2v-preview-ft-202511111122-xxxx",
        "base_model": "wan2.5-i2v-preview",
        "workspace_id": "llm-xxxxxx",
        "charge_type": "post_paid",
        "creator": "xxxxxx",
        "modifier": "xxxxxx",
        "plan": "lora"
    }
}

Verify deletion by querying status. NotFound confirms removal:

HELPCODEESCAPE-json
{
    "request_id": "eb619064-0c4f-4d29-aa49-xxxxxx",
    "message": "Not found.",
    "code": "NotFound"
}

tuning (training)-Video generation model fine-tuning API reference ​

Workflow overview ​

Before you begin ​

Supported base models ​

Upload a dataset ​

Request ​

Response ​

Example ​

Create a fine-tuning job ​

Request ​

Hyperparameters ​

Quick-start defaults ​

Full parameter reference ​

Response ​

Job status values ​

Examples ​

wan2.6-i2v ​

wan2.5-i2v-preview ​

wan2.2-i2v-flash ​

wan2.2-kf2v-flash (first and last frames) ​

Multiple training and validation sets ​

Response ​

Query job status ​

Request ​

Response ​

Example ​

Deploy a model ​

Request ​

Prompt configuration (aigc_config) ​

Response ​

Examples ​

Image-to-video (first frame) ​

Image-to-video (first and last frames) ​

Response ​

Query deployment status ​

Request ​

Response ​

Deployment status values ​

Example ​

Checkpoint management ​

List checkpoints ​

Request ​

Response ​

Example ​

Get validation details for a checkpoint ​

Request ​

Response ​

Example ​

Export a checkpoint ​

Request ​

Response ​

Example ​

Get exported model details ​

Request ​

Response ​

Export status values ​

Example ​

Deploy and call an exported checkpoint ​

Fine-tuning job management ​

Query job logs ​

List fine-tuning jobs ​

Cancel a fine-tuning job ​

Delete a fine-tuning job ​

Deployed model management ​

Delete a deployed model ​

tuning (training)-Video generation model fine-tuning API reference

Workflow overview

Before you begin

Supported base models

Upload a dataset

Request

Response

Example

Create a fine-tuning job

Request

Hyperparameters

Quick-start defaults

Full parameter reference

Response

Job status values

Examples

wan2.6-i2v

wan2.5-i2v-preview

wan2.2-i2v-flash

wan2.2-kf2v-flash (first and last frames)

Multiple training and validation sets

Response

Query job status

Request

Response

Example

Deploy a model

Request

Prompt configuration (aigc_config)

Response

Examples

Image-to-video (first frame)

Image-to-video (first and last frames)

Response

Query deployment status

Request

Response

Deployment status values

Example

Checkpoint management

List checkpoints

Request

Response

Example

Get validation details for a checkpoint

Request

Response

Example

Export a checkpoint

Request

Response

Example

Get exported model details

Request

Response

Export status values

Example

Deploy and call an exported checkpoint

Fine-tuning job management

Query job logs

List fine-tuning jobs

Cancel a fine-tuning job

Delete a fine-tuning job

Deployed model management

Delete a deployed model