Wan (万相) image prompt builder from Alibaba's own guide: its formula and terms, a separate negative prompt, per-model length check and API request.
Chinese AI Image Prompt Builder
Write a Chinese image prompt to the formula in the official Wan text-to-image guide, with the negative prompt kept separate, then check it against your model's length limit and copy the API request.
Model and API request
Aspect ratio (the size parameter)
Alibaba publishes no ratio table for these models, so these sizes are computed by this tool to its rule of 512 to 1440 pixels a side.
n is set to 1: the API makes 4 images by default and bills each one.
How to write an image prompt here
Write the subject and the scene
The subject is the main thing in the picture: say what it looks like and what it is doing. The scene is where it is.
Tap a style and the camera language
Style, shot size, angle and lens are buttons, and each one is written the way the official Wan text-to-image prompt guide writes it. Add any other style in the box beside them.
Add mood and detail
Mood words set the feeling, such as dreamy, lonely or grand. Detail is the finishing touch: where the light comes from, props, the setting, or "high resolution".
Copy each box to its own field
The prompt goes in your tool's prompt field, and the negative prompt only in its negative-prompt field.
Using the API? Pick a model and a ratio
The counter checks the prompt against the model's length limit, because Alibaba cuts off anything past it without an error. Tap a ratio and the request body below is ready to send.
Why it follows the guide's formula
This page writes prompts to Alibaba Cloud Model Studio's text-to-image prompt guide for its Wan models. The guide gives two formulas: the basic one is subject + scene + style, and the advanced one is subject (with its description) + scene + style + camera language + mood words + finishing detail. What you write and tap is joined into one description in the advanced formula's order.
The page used to join its fields under Markdown "##" headings, so the heading text was pasted into the model with the picture, and it tacked the negative terms onto the same text, while the guide treats prompt and negative_prompt as two separate parameters. The prompt is now one description, and the negative prompt has its own box.
Three things matter when you call the API, and they are in Alibaba's text-to-image API reference rather than the guide: each model has its own length limit and cuts off anything past it; the picture's dimensions are set by the size parameter, and default to 1280*1280 when it is left out; and prompt_extend rewrites your prompt unless you turn it off. The model, ratio and rewriting controls are there for those three.
Other vendors' official guides are not cited on this page, so it does not claim the same wording works equally well in Jimeng, Kling or Hunyuan.
Frequently Asked Questions
From Alibaba Cloud Model Studio's text-to-image prompt guide for its Wan models. Its advanced formula is subject (with its description) + scene + style + camera language + mood words + finishing detail, and every shot-size, angle, lens, style and lighting term on the buttons comes from the guide's term tables.
The guide describes two parameters: prompt, the positive prompt describing the picture you want, and negative_prompt, describing what you do not want to see. They are separate parameters, so you get them separately here.
Each model has a length limit: 2100 characters for wan2.6-t2i, 2000 for wan2.5-t2i-preview, 800 for wanx2.0-t2i-turbo and 500 for wan2.2 and wan2.1, and 500 for every negative prompt. Anything past it is cut off without an error, so pick your model and watch the counter.
By default, yes: prompt_extend is on, and a language model rewrites the prompt first. Untick it to use your words as written. Alibaba also notes that a rewritten prompt can trip its copyright review; if you get that error, turn rewriting off and try again.
You can paste it into any of them, but the formula and terms here are the ones Alibaba documents for Wan. Other vendors' guides are not cited on this page, so check the tool you use.
No. The prompt is put together in your browser. No model is called and nothing is sent anywhere.
Method & sources
How it computes
In the browser, by a tested kernel: subject, scene, style, camera language, mood and detail joined into one description in the order of the advanced formula in Alibaba Cloud's 万相 (Wan) text-to-image prompt guide, with the negative prompt separate. Each is counted against the chosen model's limit, and the request body is shaped as that model's API reference shapes it (input.messages for wan2.6, input.prompt below it). No model is called.
What this tool implements
- Corrected 2026-10-21: the page sent readers to ERNIE ViLG (文心一格), but yige.baidu.com now redirects to Baidu's 文心助手, whose own description lists image creation (图片创作). Every mention, in all three languages and the tool config, now names Wenxin (文心). The other four products were checked on their own sites: Jimeng (ByteDance), Wan (万相, Alibaba Tongyi; the page's older name 通义万相 is kept as a recognisable alias), Hunyuan (Tencent) and Kolors (可图, Kuaishou).
- Rebuilt 2026-11-03 in all three languages. The page joined its fields under Markdown "##" headings, so each heading was pasted into the model with the picture, and it appended the negative terms to the same text, while the 万相 guide gives prompt and negative_prompt as separate parameters. The prompt is now one description in the guide's advanced-formula order, and the negative prompt has its own box and copy button.
- Every shot-size, angle, lens and lighting term is written as the guide's prompt column writes it (特写镜头, 俯视视角), each style as its example prompt uses it (国风水墨风格, 浅水彩), and 梦幻, 孤独 and 宏伟 are its own examples of mood words. 陶瓷 is left out because its example names a subject (瓷器小狗), not a style phrase.
- Extended 2026-11-04, when /tongyi-wanxiang-prompt-builder/ was merged into this page: its model picker and length counter, size values, prompt_extend switch and API request body now sit under the prompt, written by the same tested kernel. Every figure was re-read in Alibaba's API reference on 2026-11-04 and matched what that page carried.
- The sizes for wan2.2 and below are COMPUTED: Alibaba gives those models a rule (each side 512 to 1440 pixels, default 1024*1024) and no ratio table, and the page says so beside them. The body asks for n = 1 because the API default is 4 images, each billed.
- Merged 2026-11-04: /tongyi-wanxiang-prompt-builder/ now redirects here. Its model picker and length limits, size values, prompt_extend switch and API request body were moved onto this page first, each re-read in Alibaba's API reference the same day.
Sources
- Jimeng (即梦AI), ByteDance: official site. https://jimeng.jianying.com/
- Wan (万相), Alibaba Tongyi: 「万相是阿里巴巴通义旗下的AI创意创作平台…提供文生图、图生图…」. https://tongyi.aliyun.com/wan/
- Tencent Hunyuan (腾讯混元): official site. https://hunyuan.tencent.com/
- Kolors (可图大模型), Kuaishou: official site. https://kolors.kuaishou.com/
- Baidu 文心助手: description lists 图片创作; yige.baidu.com (文心一格) redirects here. https://wenxin.baidu.com/
- Alibaba Cloud Model Studio. 文生图Prompt指南 (万相): the basic and advanced formulas, the prompt and negative_prompt parameters, and the term tables every button is taken from. https://help.aliyun.com/zh/model-studio/text-to-image-prompt — read 2026-11-03.
- Alibaba Cloud Model Studio. 文生图 V2 API reference (万相): prompt limits (wan2.6-t2i 2100, wan2.5-t2i-preview 2000, wanx2.0-t2i-turbo 800, wan2.2 and wan2.1 500), negative_prompt 500, 「超过部分会自动截断」; size rules and 「常见比例推荐的分辨率」; the wan2.6 input.messages shape; n 「默认为 4」 and billed per image; prompt_extend 「true：默认值」. https://help.aliyun.com/zh/model-studio/text-to-image-v2-api-refe… — read 2026-11-04.
Pick up where you left off
Stored only in this browser — never sent to our servers.