PUBLIC DEVELOPER API

生成平台 API 文档

第三方应用通过统一任务 API 调用所有已发布的图像、视频、语音、音乐和理解模型。模型能力与输入参数以本页实时发布契约为准。

API 版本0.2.0Base URL公开模型32

认证

除健康检查和文档目录外，所有公开调用都必须使用开发者控制台创建的 API Key。API Key 只显示给账户所有者，请勿放入浏览器前端代码或公开仓库。

HTTP Header

Authorization: Bearer <API_KEY>

管理 API Key

快速开始

1读取模型契约

先获取模型目录，使用返回的模型、变体和参数，不要在客户端写死分辨率。

2创建幂等任务

每次业务请求生成唯一的 idempotency_key，网络重试时复用同一个值。

3查询或接收结果

轮询任务详情，或在创建任务时提交公网 webhook_url。

列出模型

curl -X GET '/v1/models' \
  -H 'Authorization: Bearer <API_KEY>'

创建任务

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "image_generation",
  "model_key": "gpt-image-2",
  "variant_key": "GPT 1K 模型组",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "1k",
    "aspect_ratio": "1:1"
  }
}'

查询结果

curl -X GET '/v1/jobs/job_123456' \
  -H 'Authorization: Bearer <API_KEY>'

API 参考

这里只列出第三方集成接口。运营管理、内部执行和账户控制台路由不属于公开调用契约。

服务状态

GET/health服务健康状态无需认证

查询 API 服务名称和当前版本。

cURL

curl -X GET '/health'

GET/health/ready服务就绪状态无需认证

检查数据库等依赖是否可以接收请求。

cURL

curl -X GET '/health/ready'

文档发现

GET/v1/docs/catalog公开文档目录无需认证

读取公开端点清单和当前已发布模型契约。

cURL

curl -X GET '/v1/docs/catalog'

模型目录

GET/v1/models可调用模型目录API Key

按能力读取模型、变体、价格和输入契约。

cURL

curl -X GET '/v1/models' \
  -H 'Authorization: Bearer <API_KEY>'

GET/v1/models/cards已发布模型卡API Key

读取面向调用者展示的全部模型卡。

cURL

curl -X GET '/v1/models/cards' \
  -H 'Authorization: Bearer <API_KEY>'

GET/v1/models/{model_key}/{variant_key}单个模型契约API Key

读取指定模型变体的最新公开参数。

参数	位置	类型	必填	默认值 / 限制
`model_key`	path	string	是	-
`variant_key`	path	string	是	-

cURL

curl -X GET '/v1/models/gpt-image-2/GPT 1K 模型组' \
  -H 'Authorization: Bearer <API_KEY>'

生成任务

POST/v1/jobs创建生成任务API Key

创建幂等任务并冻结本次调用的计费报价。

参数	位置	类型	必填	默认值 / 限制
`capability_key`	body	string	是	-
`model_key`	body	string	否	-
`variant_key`	body	string	否	-
`idempotency_key`	body	string	是	1-256 chars
`input_payload`	body	object	是	256 KiB
`webhook_url`	body	public HTTP(S) URL	否	-
`metadata`	body	object	否	64 KiB

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "image_generation",
  "model_key": "gpt-image-2",
  "variant_key": "GPT 1K 模型组",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "1k",
    "aspect_ratio": "1:1"
  }
}'

GET/v1/jobs任务列表API Key

分页查询当前 API Key 所属账户的任务。

参数	位置	类型	必填	默认值 / 限制
`state`	query	queued \| running \| succeeded \| failed \| cancelled \| expired	否	-
`limit`	query	integer	否	20；1-100
`offset`	query	integer	否	0；>= 0

cURL

curl -X GET '/v1/jobs?limit=20&offset=0' \
  -H 'Authorization: Bearer <API_KEY>'

GET/v1/jobs/{job_id}任务详情API Key

读取任务状态、结果、时间和计费信息。

参数	位置	类型	必填	默认值 / 限制
`job_id`	path	string	是	-

cURL

curl -X GET '/v1/jobs/job_123456' \
  -H 'Authorization: Bearer <API_KEY>'

GET/v1/jobs/{job_id}/attempts执行记录API Key

读取任务对调用者可见的执行尝试状态。

参数	位置	类型	必填	默认值 / 限制
`job_id`	path	string	是	-

cURL

curl -X GET '/v1/jobs/job_123456/attempts' \
  -H 'Authorization: Bearer <API_KEY>'

POST/v1/jobs/{job_id}/cancel取消任务API Key

取消排队任务或请求中止正在执行的任务。

参数	位置	类型	必填	默认值 / 限制
`job_id`	path	string	是	-

cURL

curl -X POST '/v1/jobs/job_123456/cancel' \
  -H 'Authorization: Bearer <API_KEY>'

Webhook

POST/v1/webhooks/test测试 WebhookAPI Key

向公开 HTTP(S) 地址发送签名测试事件。

参数	位置	类型	必填	默认值 / 限制
`url`	body	public HTTP(S) URL	是	-
`secret`	body	string	是	16-512 chars

cURL

curl -X POST '/v1/webhooks/test' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "url": "https://example.com/webhooks/omnigen",
  "secret": "replace-with-your-signing-secret"
}'

模型参数

以下内容来自生成平台当前发布目录，包含工作流模型和外部供应商模型，不依赖内部工作流页面。

GPT 1K 模型组gpt-image-2 / GPT 1K 模型组

image_generation

类型image_generation契约版本model-card-r1价格1K 0.04 CNY

支持方式：文生图、图生图

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	必填	-
`resolution`	string	必填	1k
`aspect_ratio`	string	必填	1:1 \| 2:3 \| 3:2 \| 3:4 \| 4:3 \| 4:5 \| 5:4 \| 9:16 \| 16:9 \| 21:9
`input_image_urls`	array<string>	可选	至少 1 项；最多 16 项

分辨率与宽高比

档位	宽高比	输出尺寸
1K Fast1k	`1:1`	`1280 x 1280`
1K Fast1k	`2:3`	`848 x 1280`
1K Fast1k	`3:2`	`1280 x 848`
1K Fast1k	`3:4`	`960 x 1280`
1K Fast1k	`4:3`	`1280 x 960`
1K Fast1k	`4:5`	`1024 x 1280`
1K Fast1k	`5:4`	`1280 x 1024`
1K Fast1k	`9:16`	`720 x 1280`
1K Fast1k	`16:9`	`1280 x 720`
1K Fast1k	`21:9`	`1280 x 544`

请求 JSON

{
  "capability_key": "image_generation",
  "model_key": "gpt-image-2",
  "variant_key": "GPT 1K 模型组",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "1k",
    "aspect_ratio": "1:1"
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "image_generation",
  "model_key": "gpt-image-2",
  "variant_key": "GPT 1K 模型组",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "1k",
    "aspect_ratio": "1:1"
  }
}'

AIHub Seedance 2.0 Fast（可人脸）aihub-seedance-2.0-fast / aihub-seedance-2.0-fast

video_generation

类型video_generation契约版本model-card-r1价格720P 16.666667 CNY

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	必填	-
`resolution`	string	必填	720p
`aspect_ratio`	string	必填默认 16:9	1:1 \| 3:4 \| 4:3 \| 16:9 \| 21:9 \| 9:16
`duration_seconds`	integer	必填	15
`images`	array<string>	可选	至少 1 项；最多 4 项
`reference_videos`	array<string>	可选	至少 1 项；最多 3 项
`reference_audios`	array<string>	可选	至少 1 项；最多 1 项
`image_url`	string	可选	-

分辨率与宽高比

档位	宽高比	输出尺寸
720P720p	`16:9`	`1280 x 720`
720P720p	`9:16`	`720 x 1280`
720P720p	`1:1`	`720 x 720`

请求 JSON

{
  "capability_key": "video_generation",
  "model_key": "aihub-seedance-2.0-fast",
  "variant_key": "aihub-seedance-2.0-fast",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "720p",
    "aspect_ratio": "16:9",
    "duration_seconds": "15"
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video_generation",
  "model_key": "aihub-seedance-2.0-fast",
  "variant_key": "aihub-seedance-2.0-fast",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "720p",
    "aspect_ratio": "16:9",
    "duration_seconds": "15"
  }
}'

AIHub Seedance 2.0 满血（可人脸）aihub-seedance-2.0-full / aihub-seedance-2.0-full

video_generation

类型video_generation契约版本model-card-r1价格720P 20 CNY

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	必填	-
`resolution`	string	必填	720p
`aspect_ratio`	string	必填默认 16:9	1:1 \| 3:4 \| 4:3 \| 16:9 \| 21:9 \| 9:16
`duration_seconds`	integer	必填	15
`images`	array<string>	可选	至少 1 项；最多 4 项
`reference_videos`	array<string>	可选	至少 1 项；最多 3 项
`reference_audios`	array<string>	可选	至少 1 项；最多 1 项
`image_url`	string	可选	-

分辨率与宽高比

档位	宽高比	输出尺寸
720P720p	`16:9`	`1280 x 720`
720P720p	`9:16`	`720 x 1280`
720P720p	`1:1`	`720 x 720`

请求 JSON

{
  "capability_key": "video_generation",
  "model_key": "aihub-seedance-2.0-full",
  "variant_key": "aihub-seedance-2.0-full",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "720p",
    "aspect_ratio": "16:9",
    "duration_seconds": "15"
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video_generation",
  "model_key": "aihub-seedance-2.0-full",
  "variant_key": "aihub-seedance-2.0-full",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "720p",
    "aspect_ratio": "16:9",
    "duration_seconds": "15"
  }
}'

AIHub Seedance 2.0 Miniaihub-seedance-2.0-mini / aihub-seedance-2.0-mini

video_generation

类型video_generation契约版本model-card-r1价格720P 13.333333 CNY

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	必填	-
`resolution`	string	必填	720p
`aspect_ratio`	string	必填默认 16:9	1:1 \| 3:4 \| 4:3 \| 16:9 \| 21:9 \| 9:16
`duration_seconds`	integer	必填	15
`images`	array<string>	可选	至少 1 项；最多 4 项
`reference_videos`	array<string>	可选	至少 1 项；最多 3 项
`reference_audios`	array<string>	可选	至少 1 项；最多 1 项
`image_url`	string	可选	-

分辨率与宽高比

档位	宽高比	输出尺寸
720P720p	`16:9`	`1280 x 720`
720P720p	`9:16`	`720 x 1280`
720P720p	`1:1`	`720 x 720`

请求 JSON

{
  "capability_key": "video_generation",
  "model_key": "aihub-seedance-2.0-mini",
  "variant_key": "aihub-seedance-2.0-mini",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "720p",
    "aspect_ratio": "16:9",
    "duration_seconds": "15"
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video_generation",
  "model_key": "aihub-seedance-2.0-mini",
  "variant_key": "aihub-seedance-2.0-mini",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "720p",
    "aspect_ratio": "16:9",
    "duration_seconds": "15"
  }
}'

clmm-mg-seedance2-0-480p-miniclmm-mg-seedance2-0-480p-mini / clmm-mg-seedance2-0-480p-mini

video_generation

类型video_generation契约版本model-card-r1价格480P 0.13 CNY

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	必填	-
`resolution`	string	必填	480p
`aspect_ratio`	string	必填	16:9 \| 9:16 \| 1:1
`duration_seconds`	integer	必填	最小 5；最大 15；1 的倍数
`reference_image_urls`	array<string>	可选	至少 1 项；最多 4 项
`reference_videos`	array<string>	可选	至少 1 项；最多 3 项

分辨率与宽高比

档位	宽高比	输出尺寸
480P480p	`16:9`	`854 x 480`
480P480p	`9:16`	`480 x 854`
480P480p	`1:1`	`480 x 480`

请求 JSON

{
  "capability_key": "video_generation",
  "model_key": "clmm-mg-seedance2-0-480p-mini",
  "variant_key": "clmm-mg-seedance2-0-480p-mini",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 5
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video_generation",
  "model_key": "clmm-mg-seedance2-0-480p-mini",
  "variant_key": "clmm-mg-seedance2-0-480p-mini",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 5
  }
}'

dashscope-qwen3-omni-flashdashscope-qwen3-omni-flash / dashscope-qwen3-omni-flash

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen3-omni-flash",
  "variant_key": "dashscope-qwen3-omni-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen3-omni-flash",
  "variant_key": "dashscope-qwen3-omni-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

dashscope-qwen3.5-omni-plusdashscope-qwen3.5-omni-plus / dashscope-qwen3.5-omni-plus

chat_generation

类型chat_generation契约版本model-card-r1价格0.04 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen3.5-omni-plus",
  "variant_key": "dashscope-qwen3.5-omni-plus",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen3.5-omni-plus",
  "variant_key": "dashscope-qwen3.5-omni-plus",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

dashscope-qwen3.7-flashdashscope-qwen3.7-flash / dashscope-qwen3.7-flash

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen3.7-flash",
  "variant_key": "dashscope-qwen3.7-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen3.7-flash",
  "variant_key": "dashscope-qwen3.7-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

dashscope-qwen35-omni-flashdashscope-qwen35-omni-flash / dashscope-qwen35-omni-flash

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen35-omni-flash",
  "variant_key": "dashscope-qwen35-omni-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "dashscope-qwen35-omni-flash",
  "variant_key": "dashscope-qwen35-omni-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

deepseek-v4-flashdeepseek-v4-flash / deepseek-v4-flash

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "deepseek-v4-flash",
  "variant_key": "deepseek-v4-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "deepseek-v4-flash",
  "variant_key": "deepseek-v4-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

GPT Image 2 共享通道gpt-image-2 / gpt-image-2-shared

image_generation

类型image_generation契约版本model-card-r1价格1K 0.04 CNY · 2K 0.06 CNY

共享容量池；统一承载 1K 与 2K，请求先按分辨率过滤，再按容量权重和优先级调度。

支持方式：文生图、图生图

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	必填	-
`resolution`	string	必填	1k \| 2k
`aspect_ratio`	string	必填	1:1 \| 2:3 \| 3:2 \| 3:4 \| 4:3 \| 4:5 \| 5:4 \| 9:16 \| 16:9 \| 21:9
`quality`	string	可选	low \| medium
`input_image_urls`	array<string>	可选	至少 1 项；最多 16 项

分辨率与宽高比

档位	宽高比	输出尺寸
1K Fast1k	`1:1`	`1280 x 1280`
1K Fast1k	`2:3`	`848 x 1280`
1K Fast1k	`3:2`	`1280 x 848`
1K Fast1k	`3:4`	`960 x 1280`
1K Fast1k	`4:3`	`1280 x 960`
1K Fast1k	`4:5`	`1024 x 1280`
1K Fast1k	`5:4`	`1280 x 1024`
1K Fast1k	`9:16`	`720 x 1280`
1K Fast1k	`16:9`	`1280 x 720`
1K Fast1k	`21:9`	`1280 x 544`
2K Recommended2k	`1:1`	`2048 x 2048`
2K Recommended2k	`2:3`	`1360 x 2048`
2K Recommended2k	`3:2`	`2048 x 1360`
2K Recommended2k	`3:4`	`1536 x 2048`
2K Recommended2k	`4:3`	`2048 x 1536`
2K Recommended2k	`4:5`	`1632 x 2048`
2K Recommended2k	`5:4`	`2048 x 1632`
2K Recommended2k	`9:16`	`1152 x 2048`
2K Recommended2k	`16:9`	`2048 x 1152`
2K Recommended2k	`21:9`	`2048 x 864`

请求 JSON

{
  "capability_key": "image_generation",
  "model_key": "gpt-image-2",
  "variant_key": "gpt-image-2-shared",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "1k",
    "aspect_ratio": "1:1"
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "image_generation",
  "model_key": "gpt-image-2",
  "variant_key": "gpt-image-2-shared",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "1k",
    "aspect_ratio": "1:1"
  }
}'

metaso-minimax-h3metaso-minimax-h3 / metaso-minimax-h3

video_generation

类型video_generation契约版本model-card-r1价格768P 0.13 CNY · 2K 0.21 CNY

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	必填	-
`resolution`	string	必填	768p \| 2k
`aspect_ratio`	string	必填	adaptive \| 21:9 \| 16:9 \| 4:3 \| 1:1 \| 3:4 \| 9:16
`duration_seconds`	integer	必填	4 \| 5 \| 6 \| 7 \| 8 \| 9 \| 10 \| 11 \| 12 \| 13 \| 14 \| 15
`image_url`	string	可选	-
`end_frame_url`	string	可选	-
`reference_images`	array<string>	可选	至少 1 项；最多 9 项
`reference_videos`	array<string>	可选	至少 1 项；最多 3 项
`reference_audios`	array<string>	可选	至少 1 项；最多 3 项

分辨率与宽高比

档位	宽高比	输出尺寸
768P768p	`16:9`	`1366 x 768`
768P768p	`9:16`	`768 x 1366`
768P768p	`1:1`	`768 x 768`
2K2k	`16:9`	`2560 x 1440`
2K2k	`9:16`	`1440 x 2560`
2K2k	`1:1`	`1440 x 1440`

请求 JSON

{
  "capability_key": "video_generation",
  "model_key": "metaso-minimax-h3",
  "variant_key": "metaso-minimax-h3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "768p",
    "aspect_ratio": "adaptive",
    "duration_seconds": "4"
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video_generation",
  "model_key": "metaso-minimax-h3",
  "variant_key": "metaso-minimax-h3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "768p",
    "aspect_ratio": "adaptive",
    "duration_seconds": "4"
  }
}'

mini-minimax-m3mini-minimax-m3 / mini-minimax-m3

chat_generation

类型chat_generation契约版本model-card-r1价格0.003 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "mini-minimax-m3",
  "variant_key": "mini-minimax-m3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "mini-minimax-m3",
  "variant_key": "mini-minimax-m3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

MiniMax H3 Ref2VA V2minimax_h3_ref2va_v2 / minimax_h3_ref2va_v2

video.generate

类型video契约版本model-card-r1价格480P 0.02 CNY · 720P 0.03 CNY

Complete MiniMax H3 Ref2VA reference-to-video workflow with preview and refinement sampling.

输入参数

参数	类型	必填 / 默认	取值规则
`reference_images`	array<string>	可选	至少 0 项；最多 9 项
`reference_audio_urls`	array<string>	可选	至少 0 项；最多 2 项
`prompt`	string	必填	最多 32000 字符
`seed`	integer	可选默认 42	最小 0；最大 9223372036854776000
`resolution`	string	必填默认 480p	480p \| 720p
`aspect_ratio`	string	必填默认 16:9	16:9 \| 9:16
`duration_seconds`	number	必填	最小 4；最大 15；1 的倍数

分辨率与宽高比

档位	宽高比	输出尺寸
0.2480p	`16:9`	`608 x 352`
0.2480p	`9:16`	`352 x 608`
0.3720p	`16:9`	`736 x 416`
0.3720p	`9:16`	`416 x 736`

请求 JSON

{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref2va_v2",
  "variant_key": "minimax_h3_ref2va_v2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref2va_v2",
  "variant_key": "minimax_h3_ref2va_v2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}'

MiniMax H3 Ref2VA V2 Step 1 Previewminimax_h3_ref2va_v2 / minimax_h3_ref2va_v2_step1

video.generate

类型video契约版本model-card-r1价格480P 0.02 CNY

MiniMax H3 Ref2VA first-pass preview with cloud-uploadable latent and positive conditioning artifacts.

输入参数

参数	类型	必填 / 默认	取值规则
`reference_images`	array<string>	可选	至少 0 项；最多 9 项
`reference_audio_urls`	array<string>	可选	至少 0 项；最多 2 项
`prompt`	string	必填	最多 32000 字符
`seed`	integer	可选默认 42	最小 0；最大 9223372036854776000
`resolution`	string	必填默认 480p	480p
`aspect_ratio`	string	必填默认 16:9	16:9 \| 9:16
`duration_seconds`	number	必填	最小 4；最大 15；1 的倍数

分辨率与宽高比

档位	宽高比	输出尺寸
0.2480p	`16:9`	`608 x 352`
0.2480p	`9:16`	`352 x 608`

请求 JSON

{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref2va_v2",
  "variant_key": "minimax_h3_ref2va_v2_step1",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref2va_v2",
  "variant_key": "minimax_h3_ref2va_v2_step1",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}'

MiniMax H3 Ref2VA V2 Step 2 Refinementminimax_h3_ref2va_v2 / minimax_h3_ref2va_v2_step2

video.generate

类型video契约版本model-card-r1价格480P 0.06 CNY · 720P 0.07 CNY · 736P 0.08 CNY

输入参数

参数	类型	必填 / 默认	取值规则
`latent_url`	string	必填	最多 4096 字符
`positive_conditioning_url`	string	必填	最多 4096 字符
`upscale_factor`	string	必填默认 2x	2x \| 2.5x \| 3x
`seed`	integer	必填	最小 0；最大 9223372036854776000
`duration_seconds`	integer	必填	最小 4；最大 15

请求 JSON

{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref2va_v2",
  "variant_key": "minimax_h3_ref2va_v2_step2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "latent_url": "https://assets.example.com/input.png",
    "positive_conditioning_url": "https://assets.example.com/input.png",
    "upscale_factor": "2x",
    "seed": 0,
    "duration_seconds": 4
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref2va_v2",
  "variant_key": "minimax_h3_ref2va_v2_step2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "latent_url": "https://assets.example.com/input.png",
    "positive_conditioning_url": "https://assets.example.com/input.png",
    "upscale_factor": "2x",
    "seed": 0,
    "duration_seconds": 4
  }
}'

MiniMax H3 Reference V2minimax_h3_ref_v2 / minimax_h3_ref_v2

video.generate

类型video契约版本model-card-r1价格480P 0.1 CNY · 720P 0.1 CNY

Complete MiniMax H3 reference-to-video workflow with preview and refinement sampling.

输入参数

参数	类型	必填 / 默认	取值规则
`reference_images`	array<string>	可选	至少 0 项；最多 9 项
`reference_audio_urls`	array<string>	可选	至少 0 项；最多 2 项
`prompt`	string	必填	最多 32000 字符
`seed`	integer	可选默认 42	最小 0；最大 9223372036854776000
`resolution`	string	必填默认 480p	480p \| 720p
`aspect_ratio`	string	必填默认 16:9	16:9 \| 9:16
`duration_seconds`	number	必填	最小 4；最大 15；1 的倍数

分辨率与宽高比

档位	宽高比	输出尺寸
0.2480p	`16:9`	`608 x 352`
0.2480p	`9:16`	`352 x 608`
0.3720p	`16:9`	`736 x 416`
0.3720p	`9:16`	`416 x 736`

请求 JSON

{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref_v2",
  "variant_key": "minimax_h3_ref_v2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref_v2",
  "variant_key": "minimax_h3_ref_v2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}'

MiniMax H3 Reference V2 Step 1 Previewminimax_h3_ref_v2 / minimax_h3_ref_v2_step1

video.generate

类型video契约版本model-card-r1价格480P 0.02 CNY

MiniMax H3 fast first-pass preview with cloud-uploadable latent and positive conditioning artifacts.

输入参数

参数	类型	必填 / 默认	取值规则
`reference_images`	array<string>	可选	至少 0 项；最多 9 项
`reference_audio_urls`	array<string>	可选	至少 0 项；最多 2 项
`prompt`	string	必填	最多 32000 字符
`seed`	integer	可选默认 42	最小 0；最大 9223372036854776000
`resolution`	string	必填默认 480p	480p
`aspect_ratio`	string	必填默认 16:9	16:9 \| 9:16
`duration_seconds`	number	必填	最小 4；最大 15；1 的倍数

分辨率与宽高比

档位	宽高比	输出尺寸
0.2480p	`16:9`	`608 x 352`
0.2480p	`9:16`	`352 x 608`

请求 JSON

{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref_v2",
  "variant_key": "minimax_h3_ref_v2_step1",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref_v2",
  "variant_key": "minimax_h3_ref_v2_step1",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate.",
    "resolution": "480p",
    "aspect_ratio": "16:9",
    "duration_seconds": 4
  }
}'

MiniMax H3 Reference V2 Step 2 Refinementminimax_h3_ref_v2 / minimax_h3_ref_v2_step2

video.generate

类型video契约版本model-card-r1价格480P 0.06 CNY · 720P 0.07 CNY · 736P 0.08 CNY

MiniMax H3 refinement pass that downloads the first-pass latent and positive conditioning from cloud storage.

输入参数

参数	类型	必填 / 默认	取值规则
`latent_url`	string	必填	最多 4096 字符
`positive_conditioning_url`	string	必填	最多 4096 字符
`upscale_factor`	string	必填默认 2x	2x \| 2.5x \| 3x
`seed`	integer	必填	最小 0；最大 9223372036854776000
`duration_seconds`	integer	必填	最小 4；最大 15

请求 JSON

{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref_v2",
  "variant_key": "minimax_h3_ref_v2_step2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "latent_url": "https://assets.example.com/input.png",
    "positive_conditioning_url": "https://assets.example.com/input.png",
    "upscale_factor": "2x",
    "seed": 0,
    "duration_seconds": 4
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "video.generate",
  "model_key": "minimax_h3_ref_v2",
  "variant_key": "minimax_h3_ref_v2_step2",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "latent_url": "https://assets.example.com/input.png",
    "positive_conditioning_url": "https://assets.example.com/input.png",
    "upscale_factor": "2x",
    "seed": 0,
    "duration_seconds": 4
  }
}'

neurax-deepseek-v4-flashneurax-deepseek-v4-flash / neurax-deepseek-v4-flash

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "neurax-deepseek-v4-flash",
  "variant_key": "neurax-deepseek-v4-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "neurax-deepseek-v4-flash",
  "variant_key": "neurax-deepseek-v4-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

neurax-gemini-3.7-flashneurax-gemini-3.7-flash / neurax-gemini-3.7-flash

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "neurax-gemini-3.7-flash",
  "variant_key": "neurax-gemini-3.7-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "neurax-gemini-3.7-flash",
  "variant_key": "neurax-gemini-3.7-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

neurax-grok-4.6neurax-grok-4.6 / neurax-grok-4.6

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "neurax-grok-4.6",
  "variant_key": "neurax-grok-4.6",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "neurax-grok-4.6",
  "variant_key": "neurax-grok-4.6",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

neurax-minimax-m3neurax-minimax-m3 / neurax-minimax-m3

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "neurax-minimax-m3",
  "variant_key": "neurax-minimax-m3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "neurax-minimax-m3",
  "variant_key": "neurax-minimax-m3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

dashscope-qwen-omni-turboqwen-qwen-omni-turbo / qwen-qwen-omni-turbo

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "qwen-qwen-omni-turbo",
  "variant_key": "qwen-qwen-omni-turbo",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "qwen-qwen-omni-turbo",
  "variant_key": "qwen-qwen-omni-turbo",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

Qwen3 ASR Flash File Transcriptionqwen3-asr-flash-filetrans / qwen3-asr-flash-filetrans

audio.transcribe

类型audio契约版本model-card-r1价格0.00022 CNY / second

DashScope Qwen file transcription with dialogue and timestamp extraction.

输入参数

参数	类型	必填 / 默认	取值规则
`source_video_url`	string	必填	最多 8000 字符
`duration_seconds`	number	必填	最小 0.001；最大 43200
`channel_id`	array<integer>	可选默认 [0]	至少 1 项；最多 8 项
`enable_itn`	boolean	可选默认 false	-
`enable_words`	boolean	可选默认 true	-

请求 JSON

{
  "capability_key": "audio.transcribe",
  "model_key": "qwen3-asr-flash-filetrans",
  "variant_key": "qwen3-asr-flash-filetrans",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "source_video_url": "https://assets.example.com/input.png",
    "duration_seconds": 0.001
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "audio.transcribe",
  "model_key": "qwen3-asr-flash-filetrans",
  "variant_key": "qwen3-asr-flash-filetrans",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "source_video_url": "https://assets.example.com/input.png",
    "duration_seconds": 0.001
  }
}'

dashscope-qwen3-asr-flash-filetrans-v1qwen3-asr-flash-filetrans-v1 / qwen3-asr-flash-filetrans-v1

audio.transcribe

类型audio契约版本model-card-r1价格0.0033 CNY / second

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`source_video_url`	string	必填	最多 8000 字符
`duration_seconds`	number	必填	最小 0.001；最大 43200
`channel_id`	array<integer>	可选默认 [0]	至少 1 项；最多 8 项
`enable_itn`	boolean	可选默认 false	-
`enable_words`	boolean	可选默认 true	-

请求 JSON

{
  "capability_key": "audio.transcribe",
  "model_key": "qwen3-asr-flash-filetrans-v1",
  "variant_key": "qwen3-asr-flash-filetrans-v1",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "source_video_url": "https://assets.example.com/input.png",
    "duration_seconds": 0.001
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "audio.transcribe",
  "model_key": "qwen3-asr-flash-filetrans-v1",
  "variant_key": "qwen3-asr-flash-filetrans-v1",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "source_video_url": "https://assets.example.com/input.png",
    "duration_seconds": 0.001
  }
}'

dashscope-qwen-omni-turboroute-45ac20e4a24445a49a44d3df32f4f398 / route-45ac20e4a24445a49a44d3df32f4f398

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "route-45ac20e4a24445a49a44d3df32f4f398",
  "variant_key": "route-45ac20e4a24445a49a44d3df32f4f398",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "route-45ac20e4a24445a49a44d3df32f4f398",
  "variant_key": "route-45ac20e4a24445a49a44d3df32f4f398",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

Step 1o Turbo Visionstepfun-step-1o-turbo-vision / stepfun-step-1o-turbo-vision

chat_generation

类型chat_generation契约版本model-card-r1价格0.02 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "stepfun-step-1o-turbo-vision",
  "variant_key": "stepfun-step-1o-turbo-vision",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "stepfun-step-1o-turbo-vision",
  "variant_key": "stepfun-step-1o-turbo-vision",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

Step 3.7 Flashstepfun-step-3.7-flash / stepfun-step-3.7-flash

chat_generation

类型chat_generation契约版本model-card-r1价格0.03 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "stepfun-step-3.7-flash",
  "variant_key": "stepfun-step-3.7-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "stepfun-step-3.7-flash",
  "variant_key": "stepfun-step-3.7-flash",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

wangpeng1478 MiniMax M3wangpeng1478-minimax-m3 / wangpeng1478-minimax-m3

chat_generation

类型chat_generation契约版本model-card-r1价格0 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "wangpeng1478-minimax-m3",
  "variant_key": "wangpeng1478-minimax-m3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "wangpeng1478-minimax-m3",
  "variant_key": "wangpeng1478-minimax-m3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

x5m5x-qwen3.7-plusx5m5x-qwen3.7-plus / x5m5x-qwen3.7-plus

chat_generation

类型chat_generation契约版本model-card-r1价格0 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "x5m5x-qwen3.7-plus",
  "variant_key": "x5m5x-qwen3.7-plus",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "x5m5x-qwen3.7-plus",
  "variant_key": "x5m5x-qwen3.7-plus",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

x5m5xzh-MiniMax-M3x5m5xzh-MiniMax-M3 / x5m5xzh-MiniMax-M3

chat_generation

类型chat_generation契约版本model-card-r1价格0.01 CNY / request

Disabled external-provider variant pending model-route proof and pricing.

输入参数

参数	类型	必填 / 默认	取值规则
`prompt`	string	可选	最多 32000 字符
`messages`	array<object>	可选	至少 1 项；最多 200 项
`temperature`	number	可选	最小 0；最大 2
`top_p`	number	可选	最小 0；最大 1
`max_tokens`	integer	可选	最小 1
`max_completion_tokens`	integer	可选	最小 1
`stop`	string \| array<string>	可选	-
`seed`	integer	可选	-
`response_format`	object	可选	-
`tools`	array<object>	可选	-
`tool_choice`	string \| object	可选	-
`user`	string	可选	最多 256 字符

请求 JSON

{
  "capability_key": "chat_generation",
  "model_key": "x5m5xzh-MiniMax-M3",
  "variant_key": "x5m5xzh-MiniMax-M3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}

cURL

curl -X POST '/v1/jobs' \
  -H 'Authorization: Bearer <API_KEY>' \
  -H 'Content-Type: application/json' \
  -d '{
  "capability_key": "chat_generation",
  "model_key": "x5m5xzh-MiniMax-M3",
  "variant_key": "x5m5xzh-MiniMax-M3",
  "idempotency_key": "client-request-001",
  "input_payload": {
    "prompt": "Describe the result you want to generate."
  }
}'

Webhook 签名

平台发送事件时会附带事件 ID、事件类型、签名版本、算法、时间戳和签名。接收端应使用原始请求体与配置的 secret 校验 HMAC-SHA256，并拒绝过期时间戳和重复事件 ID。

签名请求头

X-OGP-Event-Id: evt_...
X-OGP-Event-Type: generation.job.succeeded
X-OGP-Signature-Version: v1
X-OGP-Signature-Algorithm: hmac-sha256
X-OGP-Signature-Timestamp: 1785123456
X-OGP-Signature: <hex-digest>

任务状态与错误

queuedrunningsucceededfailedcancelledexpired

400 / 422

请求格式、字段或模型参数不符合契约。

401

API Key 缺失、无效或已停用。

404

模型、任务不存在，或任务不属于当前账户。

409

模型不可用、幂等键冲突或任务状态不允许操作。

429

账户或平台并发已满，请遵循 Retry-After。

5xx

平台或上游暂时失败；先查询原任务，再决定是否重试。