Wan 2.2 Text-to-Video Spicy API

Wan 2.2 Text-to-Video Spicy is based on Alibaba Cloud's (Tongyi Wanxiang) text-to-video model, in its Spicy mode (uncensored improved version) with relaxed content restrictions compared to the standard version. It is a high-definition video generation model capable of generating 5-second videos from a text prompt, featuring optional negative prompts, 8 resolution options covering multiple aspect ratios, and an optional watermark.

On the iCreat platform, this model operates via an asynchronous two-step task flow (submit task and poll result) using the endpoint https://api.icreat.ai/v1/task/submit/aliyun/wan2-2/text-to-video-global. It supports two pricing tiers based on resolution: 480P and 1080P. Video duration is fixed at 5 seconds, meaning billing is effectively per generation.

Model Positioning

Wan 2.2 Text-to-Video Spicy is positioned as a cost-effective, uncensored text-to-video generation model for users needing rapid, affordable video outputs without strict content filters. By offering a low-priced 480P tier and a high-definition 1080P tier, it caters to both high-volume, budget-conscious generation tasks and premium quality outputs. The Spicy mode ensures relaxed content restrictions compared to the standard version, making it suitable for creative projects requiring broader expressive freedom, while complying with local laws and regulations.

Core Capabilities

Text-to-Video Generation

The model generates high-quality 5-second videos directly from textual prompts. It supports multiline text inputs, allowing users to craft detailed and nuanced descriptions to guide the video generation process accurately.

Negative Prompts

To refine the output and exclude unwanted elements, artifacts, or distortions, the model supports an optional negative prompt parameter. This allows for greater control over the final generated video, ensuring the content aligns closely with the user's creative vision.

Multi-Resolution Support

The model offers 8 distinct resolution options covering 16:9, 9:16, 1:1, 4:3, and 3:4 aspect ratios. These are divided into a 1080P tier for high-definition outputs and a 480P tier for cost-effective, rapid generation, providing flexibility for various display formats and platforms.

Uncensored Spicy Mode

As a Spicy mode endpoint, this model is the uncensored improved version of the standard Wan 2.2 text-to-video model. It maintains the original generation capabilities while significantly relaxing content restrictions, enabling a wider range of creative and expressive content generation.

Pricing

Total Cost = Unit Price × Output Video Duration. Note: The 8 size options map to two billing tiers — 1080P tier (19201080, 10801920, 14401440, 16321224, 12241632) and 480P tier (832480, 480832, 624624). Duration is fixed at 5 seconds, so billing is effectively per generation: $0.10 per 480P video, $0.50 per 1080P video.

Application Scenarios

• Rapid prototyping of creative video concepts without strict content filtering
• High-volume generation of short-form video content for social media at a low cost
• Generating high-definition 1080P video assets for marketing and advertising
• Creating diverse visual content across various aspect ratios (landscape, portrait, square)
• Artistic and expressive content generation requiring relaxed content restrictions

Model Comparison

Comparison Table 1: Wan 2.2 Text-to-Video Spicy vs Wan 2.6 Text-to-Video Spicy

Note: Wan 2.2 is cheaper (480P at only $0.02/s) but duration is fixed at 5 seconds with no 720P tier. Wan 2.6 offers more flexible duration and resolution options.

Comparison Table 2: Low-Price Text-to-Video Model Comparison

Note: Wan 2.2's 480P tier is the lowest price in its class, but duration is fixed at 5 seconds.

Why Choose Wan 2.2 Text-to-Video Spicy?

• Uncensored Generation: Spicy mode provides an uncensored improved version with relaxed content restrictions compared to the standard version, maximizing creative freedom.
• Cost-Effective: Offers a highly affordable 480P tier at just $0.02 per second, making high-volume video generation economical.
• High Definition: Includes a 1080P tier for crisp, high-quality video outputs suitable for professional use.
• Flexible Aspect Ratios: Supports 8 size options covering all major aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4) for diverse platform requirements.
• Negative Prompt Control: Allows the exclusion of unwanted elements, ensuring the generated video closely matches the user's intent.

Specifications

Architecture

The model utilizes an asynchronous two-step API flow. In the first step, a task is submitted to the endpoint https://api.icreat.ai/v1/task/submit/aliyun/wan2-2/text-to-video-global with the request body containing two sibling top-level objects: input (holding the prompt and optional negative_prompt) and parameters (holding the size and optional watermark). Upon successful submission, a task_id is returned. In the second step, the user polls the task result endpoint https://api.icreat.ai/v1/task/result using the same task_id. While the video is processing, the result field will be an empty array []. Once the status becomes SUCCEEDED, the result array will contain a video resource object with url and download_url, along with a costUSD field indicating the total cost of the generation.

Notes

• Use the same task_id across both the submit and result polling steps; result is [] while processing — keep polling.
• FAILED is a terminal status; check the request parameters if encountered.
• input and parameters are sibling top-level fields; do not nest them.
• Video duration is fixed at 5 seconds; the duration parameter is not supported; do not pass duration, prompt_extend, or seed.
• The size parameter has 8 supported values: 1080P tier (19201080, 10801920, 14401440, 16321224, 12241632) and 480P tier (832480, 480832, 624624).
• The prompt is recommended to be within 800 characters; negative_prompt has a maximum of 500 characters.
• The watermark parameter defaults to false and, when enabled, adds a watermark to the bottom-right corner of the video.
• Spicy mode is the uncensored improved version with relaxed content restrictions compared to the standard version; comply with local laws and regulations when using.

Frequently Asked Questions

How do I call Wan 2.2 Text-to-Video Spicy?

You call the model using a two-step asynchronous flow. First, submit a task to the endpoint https://api.icreat.ai/v1/task/submit/aliyun/wan2-2/text-to-video-global using Bearer authentication. The request body must contain two sibling top-level objects: input (for prompt and negative_prompt) and parameters (for size and watermark). After receiving the task_id, poll the https://api.icreat.ai/v1/task/result endpoint with the same task_id until the status is SUCCEEDED.

How is billing calculated?

Total Cost = Unit Price × Output Video Duration. The duration is fixed at 5 seconds, so billing is effectively per generation. A 480P video costs $0.10 per generation ($0.02/s × 5s), and a 1080P video costs $0.50 per generation ($0.10/s × 5s). The response will include a costUSD field indicating the exact amount charged upon successful generation.

What values does size support?

The size parameter supports 8 options across two tiers. The 1080P tier includes: 19201080, 10801920, 14401440, 16321224, and 12241632. The 480P tier includes: 832480, 480832, and 624624. These cover 16:9, 9:16, 1:1, 4:3, and 3:4 aspect ratios.

Can I adjust the duration?

No. The video duration is fixed at 5 seconds. The duration parameter is not supported. Do not pass duration, prompt_extend, or seed in your request parameters.

What is the difference between Spicy mode and the standard version?

Spicy mode is the uncensored improved version of the model with relaxed content restrictions compared to the standard version. It maintains the same generation capabilities but allows for a broader range of content. Users must comply with local laws and regulations when using this mode.

What is the difference from Wan 2.6 Text-to-Video Spicy?

Wan 2.2 is cheaper (offering a 480P tier at $0.02/s) but has a fixed 5-second duration and no 720P tier. Wan 2.6 supports more flexible durations of 5, 10, and 15 seconds, and offers both 720P and 1080P tiers, but at higher price points.