Qwen Image API
阿里巴巴 Qwen-Image——20B MMDiT 新一代文生图和编辑工具集,支持中英双语,具备多图编辑、LoRA 定制、分层合成和 96 种姿态的镜头角度系统。
文生图提供基础版、增强的 2512 和 2.0-pro 变体。编辑端点包括 Edit、Edit-Plus(多图、ControlNet)、Edit-LoRA、Edit-Multiple-Angles(96 姿态镜头系统)和 Layered(提示词引导的分解)。Qwen Image 2.0 系列变体也包含在同一前缀下。
/filters:quality(82)/examples/e26b4143e025f71d55e3e5704c446d2b/1783682236366122456_NdT2cmvE.webp)
概览
关于 Qwen Image API
Qwen Image 能做什么、它在 Alibaba 模型阵容中的定位,以及团队选择它的原因。
Qwen Image 是 Alibaba 推出的图像生成与编辑模型,可通过 WaveSpeedAI REST API 使用。阿里巴巴 Qwen-Image——20B MMDiT 新一代文生图和编辑工具集,支持中英双语,具备多图编辑、LoRA 定制、分层合成和 96 种姿态的镜头角度系统。
文生图提供基础版、增强的 2512 和 2.0-pro 变体。编辑端点包括 Edit、Edit-Plus(多图、ControlNet)、Edit-LoRA、Edit-Multiple-Angles(96 姿态镜头系统)和 Layered(提示词引导的分解)。Qwen Image 2.0 系列变体也包含在同一前缀下。
WaveSpeedAI 上的 Qwen Image 系列提供 21 个 REST 端点,涵盖 Image-To-Image, Text-To-Image, Training 个工作流。每个变体都有各自的定价、参数选项和示例输出——请选择与你的输入模态和生产约束相匹配的那一个,或使用同一个 API 密钥调用多个变体,组合成多步骤流水线。
使用与 WaveSpeedAI 上其他 1,000 多个 AI 模型相同的 API 密钥、账单账户和速率限制来运行 Qwen Image。无需单独对接供应商,无需各家 SDK,也无需应对各家不同的速率限制——一次集成即可覆盖从文生图、文生视频到音频合成、3D 生成、放大和编辑的全部能力。
端点
全部 Qwen Image API 端点
WaveSpeedAI 现已提供 21 个 Qwen Image 端点——请选择适合你工作流的变体。
/filters:quality(82)/media/images/20260408110013_8als1nwx.webp)
Qwen Image 2.0 Edit
Qwen Image 2.0 Edit is an advanced image-editing model with improved quality and better understanding of instructions. Up to 2k. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110922_079iorac.webp)
Qwen Image 2.0 Pro Edit
Qwen Image 2.0 Pro Edit is a professional-grade image editing model with superior quality and advanced instruction understanding. Up to 2k. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110019_rr2dv7f5.webp)
Qwen Image 2.0 Text To Image
Qwen Image 2.0 is an advanced text-to-image model with enhanced image quality and improved prompt understanding. Up to 2k. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110930_uej3wr3b.webp)
Qwen Image 2.0 Pro Text To Image
Qwen Image 2.0 Pro is a professional-grade text-to-image model with superior quality and advanced prompt understanding. Up to 2k. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110201_c4uo1ban.webp)
Qwen Image Edit 2509 Multiple Angles
Qwen Image Edit 2509 Multiple Angles is an AI image editing model that generates multiple-angle views of objects or scenes from a single image. Transform perspectives and create diverse viewpoints with text prompts. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408105730_1fu19xjx.webp)
Qwen Image Max Edit
Qwen Image Max Edit is an AI model for image editing with text prompts, supporting both Chinese and English languages. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408105738_sf8u0u8s.webp)
Qwen Image Max Text To Image
Qwen Image Max is a text-to-image model with high-quality image generation supporting Chinese and English prompts. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110235_id9bp4jx.webp)
Qwen Image Edit Multiple Angles
Generate specific camera angles from a single image using a 96-pose camera system. Control horizontal rotation, vertical tilt, and zoom to create front, side, back views and more. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408111145_9y1tzzdn.webp)
Qwen Image 2512 Lora Trainer
Qwen-Image-2512 LoRA Trainer lets you train custom LoRA models 10x faster with style, character, and object training. From concept to model in minutes, not hours—upload a ZIP file containing images to start. Ready-to-use REST inference API, best performance, no cold starts, affordable pricing.
/filters:quality(82)/media/images/20260408110225_tzvu0c8q.webp)
Qwen Image Text To Image 2512 Lora
Qwen-Image-2512 LoRA is an enhanced 20B MMDiT text-to-image model with LoRA support for fast customization and refined image generation. Ready-to-use REST inference API, best performance, no cold starts, affordable pricing.
/filters:quality(82)/media/images/20260408110216_pnfbnar4.webp)
Qwen Image Text To Image 2512
Qwen Image 2512 is Qwen's latest text-to-image model with enhanced prompt understanding, superior text rendering, and versatile aspect ratio support. Ready-to-use REST inference API, best performance, no cold starts, affordable pricing.
/filters:quality(82)/media/images/20260408110130_3mxvct8p.webp)
Qwen Image Edit 2511 Lora
Qwen Image Edit 2511 LoRA is an enhanced version with custom LoRA support for personalized styles. It delivers stronger edit consistency, robust multi-person identity/pose consistency, custom LoRA styles, enhanced industrial/product design, and improved geometric reasoning for structure-preserving edits. Built for stable production use with a ready-to-use REST API, no cold starts, and predictable pricing.
/filters:quality(82)/media/images/20260408110155_wmxhc3xt.webp)
Qwen Image Edit 2511
Qwen Image Edit 2511 is a major upgrade over 2509 for real-world image editing and design. It delivers stronger edit consistency, robust multi-person identity/pose consistency, built-in LoRA styles, enhanced industrial/product design, and improved geometric reasoning for structure-preserving edits. Built for stable production use with a ready-to-use REST API, no cold starts, and predictable pricing.
/filters:quality(82)/media/images/20260408110230_htadxp5h.webp)
Qwen Image Layered
Qwen-Image Layered is a unified image-layer decomposition model for prompt-guided compositing. Provide points, boxes, or rough masks to isolate subjects and regions, and the model splits a single image into multiple RGBA layers with clean alpha, soft edges, and correct occlusion order. Ready-to-use REST inference API with fast response, no cold starts, and affordable pricing.
/filters:quality(82)/media/images/20260408110212_z5lpi1m1.webp)
Qwen Image Edit Plus Lora
Qwen-Image-Edit-Plus (2509) is 20B MMDiT image-to-image editor supporting multi-image edits, single-image consistency, and native ControlNet. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110146_inh26a0l.webp)
Qwen Image Edit Plus
Qwen-Image-Edit-Plus (2509) is a 20B MMDiT image editor with multi-image editing, single-image consistency and native ControlNet support. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110141_4lrnn2jl.webp)
Qwen Image Edit Lora
Qwen-Image-Edit LoRA (20B) enables bilingual Chinese/English image-to-image editing with style preservation and semantic and appearance edits. Ready-to-use REST API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110150_zif3ayyb.webp)
Qwen Image Edit
Qwen-Image-Edit is a 20B MMDiT image-to-image model offering precise bilingual (Chinese & English) text edits while preserving style. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408111422_uw2jz93m.webp)
Qwen Image Lora Trainer
Train custom Qwen-Image LoRA models 10x faster. Style training, character training, object training. From concept to model in minutes, not hours. Upload a ZIP file containing images to start!
/filters:quality(82)/media/images/20260408110137_50swwx4z.webp)
Qwen Image Text To Image Lora
Qwen-Image LoRA is a 20B MMDiT next-gen text-to-image model with LoRA support for fast customization and refined image generation. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
/filters:quality(82)/media/images/20260408110208_kl4q8jw4.webp)
Qwen Image Text To Image
Qwen-Image is a 20B MMDiT next-gen text-to-image model that generates images from text prompts. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.
示例
看看 Qwen Image 的实际效果
由 Qwen Image API 生成的真实输出。悬停在任意视频上即可预览,点击可打开全尺寸查看器。
使用方法
如何使用 Qwen Image API
从注册到完成一次生成,只需四步。完整的 Python、Node.js 和 cURL 示例见下方的 API 部分。
- 01
获取 API 密钥
注册 WaveSpeedAI 账号,并从控制台复制你的 API 密钥。新账号附带免费体验额度——足够在开始计费前把 Playground 运行几十次。
- 02
提交预测
把你的输入以 JSON 形式 POST 到 https://api.wavespeed.ai/api/v3/wavespeed-ai/qwen-image/text-to-image。端点会立即返回预测 ID——生成是异步的,因此推理期间你无需保持连接。
- 03
轮询完成状态
GET https://api.wavespeed.ai/api/v3/predictions/{request_id}/result。状态为 completed 时返回输出;状态为 failed、cancelled、timeout 或 deleted 时以错误终止;其他任何状态都继续轮询。
- 04
读取输出 URL
状态变为 "completed" 后,从 data.outputs[0] 读取 URL。该 URL 指向 WaveSpeedAI CDN 上你生成的媒体——具体是图片、视频、音频还是 3D 文件,取决于你调用的 Qwen Image 变体。
应用场景
用 Qwen Image 可以构建什么
开发者和创作者使用 Qwen Image API 的常见工作流。
20B MMDiT 文生图
wavespeed-ai/qwen-image/text-to-image 是 20B MMDiT 新一代文生图模型,可根据文本提示词生成图像——是 Qwen Image 系列的基础生成端点。
文字渲染出众的增强版 2512
wavespeed-ai/qwen-image/text-to-image-2512 是 Qwen 最新的文生图模型,据目录介绍,提示词理解增强、文字渲染出众,并支持多种画幅比例。
用 Edit-Plus 进行多图编辑
wavespeed-ai/qwen-image/edit-plus 是 20B MMDiT 编辑器,支持多图编辑、单图一致性和原生 ControlNet——适合需要参考多张源图像的复杂编辑。
96 种姿态的镜头角度控制
wavespeed-ai/qwen-image/edit-multiple-angles 利用 96 种姿态的镜头系统,从单张图像生成特定的镜头角度——控制水平旋转、垂直倾斜和缩放,得到正面、侧面、背面等视图。
分层合成分解
wavespeed-ai/qwen-image/layered 是统一的图像分层分解模型,用于提示词引导的合成——提供点、框或粗略蒙版来分离主体,把单张图像拆分为多个图层。
LoRA 定制
支持 LoRA 的变体(text-to-image-2512-lora、edit-lora、edit-plus-lora)可实现快速定制和更精细的生成——训练或应用 LoRA 检查点,以保持风格、角色或品牌的一致性。
技巧
Qwen Image 提示词技巧
让 Qwen Image 输出更好结果的实用建议——总结自生产流水线中各类图像模型都适用的做法。
- 01
用 2512 获得最新的文字渲染
text-to-image-2512 是增强变体,文字渲染和提示词理解更出色——排版重要时,请选它而不是基础的 text-to-image。
- 02
用 Edit-Multiple-Angles 获得产品多角度视图
edit-multiple-angles 可根据单张产品照片生成正面、侧面、背面和自定义的镜头角度——适合电商目录,无需多机位拍摄。
- 03
多图参考使用 Edit-Plus
当编辑依赖多张源图像时,请使用支持原生 ControlNet 的 edit-plus,而不是单图编辑。
- 04
合成工作流使用 Layered
使用 layered 把图像分解为提示词引导的图层——用点、框或粗略蒙版分离主体,用于下游合成。
- 05
用 LoRA 变体保持品牌一致
通过 text-to-image-2512-lora 或 edit-plus-lora 应用训练好的 LoRA 检查点,在多次生成中保持固定的风格、角色或品牌身份。
- 06
中英双语编辑
Edit 和 Edit-LoRA 支持中英双语的图生图编辑并保留风格——适合本地化工作流。
定价
Qwen Image API 定价
按输出计费。最终费用会随你在各变体 Playground 中设置的参数(分辨率、时长、输出数量、参考素材)而变化。
API
调用 Qwen Image API
在 wavespeed.ai/accesskey 注册并获取 API 密钥,然后通过 REST 提交预测。Playground 可以为任意输入组合生成可直接粘贴的示例代码。
POSThttps://api.wavespeed.ai/api/v3/wavespeed-ai/qwen-image/text-to-image
# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
-X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/qwen-image/text-to-image" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $WAVESPEED_API_KEY" \
-d '{
"prompt": "A cinematic shot of a city at sunset, soft golden light",
"size": "1024*1024",
"output_format": "jpeg"
}')
TASK=$(printf '%s' "$SUBMIT_RESPONSE" | jq 'if has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "$TASK" | jq -r '.id')
if [ -z "$PREDICTION_ID" ] || [ "$PREDICTION_ID" = "null" ]; then
printf 'Submission response did not contain a prediction id
' >&2
exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/$PREDICTION_ID/result"
# 2. Poll until the prediction finishes.
while true; do
RESPONSE=$(curl --silent --show-error --fail-with-body "$RESULT_URL" \
-H "Authorization: Bearer $WAVESPEED_API_KEY")
RESULT=$(printf '%s' "$RESPONSE" | jq 'if has("data") then .data else . end')
STATUS=$(printf '%s' "$RESULT" | jq -r '.status')
case "$STATUS" in
completed) printf '%s\n' "$RESULT" | jq '.outputs'; break ;;
failed|cancelled|timeout|deleted) printf '%s\n' "$RESULT" | jq . >&2; exit 1 ;;
*) sleep 2 ;;
esac
doneconst submitUrl = "https://api.wavespeed.ai/api/v3/wavespeed-ai/qwen-image/text-to-image";
const apiKey = process.env.WAVESPEED_API_KEY;
if (!apiKey) throw new Error('Set WAVESPEED_API_KEY');
async function requestJson(url, options = {}) {
const response = await fetch(url, options);
if (!response.ok) throw new Error(await response.text());
return response.json();
}
// 1. Submit the prediction.
const body = await requestJson(submitUrl, {
method: "POST",
headers: {
"Authorization": `Bearer ${apiKey}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"prompt": "A cinematic shot of a city at sunset, soft golden light",
"size": "1024*1024",
"output_format": "jpeg"
}),
});
const task = body.data ?? body;
const resultUrl = `https://api.wavespeed.ai/api/v3/predictions/${task.id}/result`;
// 2. Poll until the prediction finishes.
while (true) {
const resultBody = await requestJson(resultUrl, {
headers: { "Authorization": `Bearer ${apiKey}` },
});
const result = resultBody.data ?? resultBody;
if (result.status === "completed") {
console.log(result.outputs);
break;
}
if (["failed", "cancelled", "timeout", "deleted"].includes(result.status)) throw new Error(JSON.stringify(result));
await new Promise(resolve => setTimeout(resolve, 2000));
}import json
import os
import time
from urllib.request import Request, urlopen
api_key = os.environ["WAVESPEED_API_KEY"]
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
payload = {
"prompt": "A cinematic shot of a city at sunset, soft golden light",
"size": "1024*1024",
"output_format": "jpeg"
}
def request_json(url, data=None):
request = Request(url, data=data, headers=headers, method="POST" if data else "GET")
with urlopen(request) as response:
return json.load(response)
# 1. Submit the prediction.
body = request_json("https://api.wavespeed.ai/api/v3/wavespeed-ai/qwen-image/text-to-image", json.dumps(payload).encode())
task = body.get("data", body)
result_url = f"https://api.wavespeed.ai/api/v3/predictions/{task['id']}/result"
# 2. Poll until the prediction finishes.
while True:
result_body = request_json(result_url)
result = result_body.get("data", result_body)
status = result.get("status")
if status == "completed":
print(result.get("outputs", []))
break
if status in {"failed", "cancelled", "timeout", "deleted"}:
raise RuntimeError(result)
time.sleep(2)对比
Qwen Image 与其他方案对比
在 WaveSpeedAI 上,何时应选择 Qwen Image 而不是同类模型。
Qwen Image 对比 Seedream 4.5
Seedream 4.5 强调排版,并提供用于多图身份锁定的 Sequential 变体。Qwen Image 的编辑范围更广——Edit-Plus、多角度、分层合成,以及中英双语编辑。
Qwen Image 对比 GPT Image 2
GPT Image 2 有明确的质量档位和参考图像编辑工作流。Qwen Image 提供原生 ControlNet、96 种姿态的镜头角度和分层分解——编辑方式不同,单次调用成本更低。
Qwen Image 对比 Nano Banana 2
Nano Banana 2 提供多角色一致性(最多 5 个)和网页搜索增强。Qwen Image 的优势在于编辑深度——多图 Edit-Plus、镜头角度生成和提示词引导的图层分解。
常见问题
Qwen Image API — 常见问题
定价、许可、集成——关于在 WaveSpeedAI 上运行 Qwen Image 的常见问题。
Qwen Image API 是什么?
Qwen Image 是 Alibaba 的图像生成模型,在 WaveSpeedAI 上以 REST API 形式提供。阿里巴巴 Qwen-Image——20B MMDiT 新一代文生图和编辑工具集,支持中英双语,具备多图编辑、LoRA 定制、分层合成和 96 种姿态的镜头角度系统。你可以通过编程方式调用它,也可以在上方链接的 Playground 中试用。
如何调用 Qwen Image API?
注册 WaveSpeedAI 账号,从 /accesskey 复制你的 API 密钥,然后把输入以 JSON 形式 POST 到 https://api.wavespeed.ai/api/v3/wavespeed-ai/qwen-image/text-to-image。端点会返回预测 ID。从大约每 2 秒一次开始轮询结果端点,长耗时任务可适当拉长间隔,并在任何终止状态时停止。上方有面向生产环境的 Python / Node.js / cURL 示例。
Qwen Image API 的费用是多少?
Qwen Image 每次调用 $0.02 起。实际费用会随你设置的参数(分辨率、时长、输出数量、参考素材)而变化。Playground 中“生成”按钮旁的实时费用预览会显示你当前输入对应的准确价格。
有哪些 Qwen Image 变体可用?
WaveSpeedAI 托管了 21 个已上线的 Qwen Image 端点:wavespeed-ai/qwen-image-2.0/edit, wavespeed-ai/qwen-image-2.0-pro/edit, wavespeed-ai/qwen-image-2.0/text-to-image, wavespeed-ai/qwen-image-2.0-pro/text-to-image, wavespeed-ai/qwen-image/edit-2509-multiple-angles, wavespeed-ai/qwen-image-max/edit, wavespeed-ai/qwen-image-max/text-to-image, wavespeed-ai/qwen-image/edit-multiple-angles等。每个变体都有自己的 Playground 页面和定价。
Qwen Image 的输出可以商用吗?
商用权利遵循 Alibaba 的模型许可。大多数 Alibaba 模型允许商用输出;具体许可摘要请查看各模型的 Playground 页面,平台层面的条件请参阅 WaveSpeedAI 的服务条款。
为什么要在 WaveSpeedAI 上使用 Qwen Image,而不是直接对接?
一个 API 密钥、一个账单账户,即可使用 Qwen Image 以及来自其他提供方的 1,000 多个 AI 模型。无需逐家配置 SDK,无需应对各自独立的速率限制,也无需为每家重写集成代码。价格通常与 Alibaba 直接提供的 API 持平或更低。
提供方
关于 Alibaba
Qwen Image 及 WaveSpeedAI 上 Alibaba 更多模型背后的团队。
阿里巴巴通义实验室推出了 Wan 系列视频模型和 Qwen 系列大语言模型。Wan 以开放权重发布、变体覆盖广泛(文生视频、图生视频、参考生视频、视频编辑、视频延长、图像编辑、文生图)而著称,并在运动稳定性和跨多语言提示词的提示词遵循方面表现稳定。
在 WaveSpeedAI 上用 Qwen Image 开始构建
注册即送免费体验额度。一个 API 密钥,即可使用来自 Alibaba 及其他所有提供方的 1,000 多个 AI 模型。
/filters:quality(82)/examples/ab6c113bc342972e2bf6fa3a3dd02d44/1783682238198695915_4tkuDMV6.webp)
/filters:quality(82)/examples/ae55efa5dc15feeab655c08c99c0fb67/1783682240218864185_xajsBKir.webp)
/filters:quality(82)/examples/2aae35cf0e4f367122dc3f8306577ba6/1783682242738489413_BobmvEaj.webp)
/filters:quality(82)/examples/fb556f2b9aa99c70a63a0a144d14bfe1/1783682244605044938_9FNX6enx.webp)