How do I call Alibaba Qwen models?

Create an API key on XiaoZhan AIhub, point your SDK's base_url at https://api.dflop.top/v1 and put the model ID from the list above in the model field. The OpenAI, Anthropic and Gemini SDKs all work.

How are Alibaba Qwen models billed?

Each model is billed its own way: text models per input and output token, image models per image, video models per second or per token. Prices are in credits (60 credits = ¥1); each model page lists the full price.

Alibaba Qwen models: APIs and prices

9 models·Filter in the catalogue

Qwen models for text, vision and more. XiaoZhan AIhub offers 9 Alibaba Qwen models (Video generation, Image generation, Text generation). Call them with one API key through the OpenAI, Anthropic or Gemini SDK; prices are in credits (60 credits = ¥1) and every model page lists the full price.

Text generation2

Qwen3.8-Max-cjxqwen3.8-max-cjx
Alibaba Qwen3.8-Max (cjx line). 1M context, thinking on by default, image input and tool calling, billed at Alibaba Cloud Model Studio's list price. OpenAI Chat and Responses protocols.
Input 0.504Credits/K0.72Credits·Output 1.512Credits/K2.16Credits
Qwen3.8-Omni-Flashqwen3.8-omni-flash
Alibaba Qwen3.8-Omni-Flash: the natively omni-modal Flash tier with 1M context. Takes text, images, audio and video directly at one flat price, with a thinking chain.
Input 0.048Credits/K·Output 0.162Credits/K

Image generation2

Qwen Image 3.0qwen-image-3.0
Alibaba Qwen Image 3.0 image generation (synchronous). Text-to-image and image-to-image (reference-image editing), any size from 512×512 to 2560×2560, up to 6 images per request. 1K and 2K cost the same.
per image 10.8Credits·Input image 1.2Credits/image
Qwen Image 3.0 Proqwen-image-3.0-pro
Alibaba Qwen Image 3.0 Pro image generation (synchronous), stronger at complex layouts and text rendering. Text-to-image and image-to-image, up to 6 images per request, priced in two tiers (1K / 2K) by total output pixels.
per image 15Credits–30Credits·Input image 1.2Credits/image

Video generation5

Happy Horse 1.1 Image-to-Videohappyhorse-1.1-i2v
Happy Horse 1.1 (Wan family) first-frame image-to-video (asynchronous task): 3-15 seconds, 720P / 1080P.
per second 54Creditsfrom
Happy Horse 1.1 All-round Referencehappyhorse-1.1-r2v
Happy Horse 1.1 (Wan family) multi-reference image-to-video (asynchronous task): 1-9 reference images, referred to as character1/2/3 in the prompt, 3-15 seconds, 720P / 1080P.
per second 54Creditsfrom
Happy Horse 1.1 Text-to-Videohappyhorse-1.1-t2v
Happy Horse 1.1 (Wan family) text-to-video (asynchronous task): 3-15 seconds, 720P / 1080P, 16:9 / 9:16 / 1:1 / 4:3 / 3:4 / 4:5 / 5:4 / 21:9 / 9:21. With a first-frame image it becomes image-to-video automatically (same price). Smoother motion, steadier subjects and more realistic texture than 1.0.
per second 54Creditsfrom
Wan 3.0 Videowan3.0-video
Alibaba Wan 3.0 video generation (asynchronous task): 2-30 seconds, 480P / 720P / 1080P, 30 fps, 16:9 / 9:16 / 1:1 / 4:3 / 3:4 / adaptive. One card, two modes: with a first-frame reference image it becomes image-to-video automatically (same price).
per second 18Creditsfrom
Wan 3.0 Video Primewan3.0-video-prime
Alibaba Wan 3.0 video generation, Prime accelerated version (asynchronous task): the same capabilities as the standard version with noticeably faster output. 2-30 seconds, 480P / 720P / 1080P, 30 fps, 16:9 / 9:16 / 1:1 / 4:3 / 3:4 / adaptive.
per second 27Creditsfrom

FAQ

How do I call Alibaba Qwen models?: Create an API key on XiaoZhan AIhub, point your SDK's base_url at https://api.dflop.top/v1 and put the model ID from the list above in the model field. The OpenAI, Anthropic and Gemini SDKs all work.
How are Alibaba Qwen models billed?: Each model is billed its own way: text models per input and output token, image models per image, video models per second or per token. Prices are in credits (60 credits = ¥1); each model page lists the full price.