New: GPT Image 2.6 is now live on KIE →
Wan 2.6
wan / 2-6-image-to-video
Commercial usePart of the Wan API family — compare all Wan models →

Wan 2.6 is Alibaba’s latest AI video model, offering affordable multi-shot 1080p generation with stable characters and synchronized native audio. Through the Wan 2.6 API—including T2V, I2V, and reference-guided modes—you can create up to 15-second cinematic videos with improved motion logic, consistent visuals, and production-ready quality.

Model Type:
Pricing: 70 / 140 / 210 credits (~$0.35 / $0.70 / $1.05) for 5 / 10 / 15 s at 720 p and 105 / 210 / 315 credits (~$0.53 / $1.05 / $1.58) for 5 / 10 / 15 s at 1080 p — flat, all-inclusive rates roughly 30 % cheaper than Fal. High-tier top-ups (+10% bonus) bring effective pricing down to ~$0.318 / ~$0.636 / ~$0.955 (720 p 5 / 10 / 15 s) and ~$0.482 / ~$0.955 / ~$1.436 (1080 p 5 / 10 / 15 s).
Input

Text prompts for video generation. Supports both Chinese and English, with a minimum of 2 characters and a maximum of 5,000 characters.

0/5000

Click to upload or drag and drop

Supported formats: JPEG, PNG, WEBP Maximum file size: 10MB; Maximum files: 1

A list of image URLs. All images must be at least 256x256px.

The duration of the generated video in seconds

Video resolution tier

The multi shots parameter controls the shot composition style during AI video generation, determining whether the generated video is a single continuous shot or multiple shots with transitions.

A configurable parameter. Defaults to true in the Playground.

There is no guarantee that everything can be filtered out; if you are not satisfied with the results, you will need to make your own arrangements.
Output
output typevideo
Examples

Explore different use cases and parameter configurations

README

Complete guide to using Wan 2.6

Affordable Wan 2.6 API for Cinematic Multi-Shot Video Generation

Cinematic multi-shot video generation at an affordable price—featuring rich backgrounds, stable characters, and synchronized native audio, powered by Alibaba’s Wan 2.6 API.

Original image

Introducing Wan 2.6 API: A New Generation of AI Video Creation from Wan AI

Developed by Wan AI within the Alibaba ecosystem, Wan 2.6 is the latest generation of Wan’s AI video model, focused on turning short prompts and visual inputs into coherent, multi-shot video stories. Version 2.6 introduces stronger scene continuity, more stable characters, and improved control over camera movement and pacing, making generated videos feel deliberate rather than fragmented.Built on top of the Wan 2.6 model, Wan 2.6 API makes these capabilities available for real-world products and workflows. It allows developers to generate videos from text, images, or reference videos, supporting Text-to-Video, Image-to-Video, and Reference-to-Video creation with up to 15-second HD output, native audio, and accurate lip sync for commercial use.

Video Generation Modes in Wan 2.6 API on Kie.ai

Wan 2.6 T2V API (Text To Video)

Wan 2.6 T2V API generates cinematic videos directly from natural language. Wan 2.6 understands multi-shot prompts and storyboard-style descriptions, translating shot order, camera direction, pacing, and mood into a coherent video sequence rather than a single isolated clip. This mode is well suited for scripts, briefs, and structured scene descriptions.

Wan 2.6 I2V API (Image To Video)

Wan 2.6 I2V API animates a single image into motion while preserving subject identity and visual style. Wan 2.6 maintains facial features, proportions, textures, and overall composition, making it suitable for portraits, product images, illustrations, and other static visuals that need to be extended into short-form video.

Wan 2.6 R2V API (Reference to Video)

Wan 2.6 Reference to Video API allows you to use an uploaded reference video to guide the generation of new scenes. Wan 2.6 extracts key visual characteristics—such as appearance, style, and voice—from the reference and applies them consistently to newly generated videos, enabling character continuity across shots and related content.

What’s New in Wan 2.6 API for AI Video Generation

Multi-Shot Storytelling with Cinematic Precision using Wan 2.6 API

The Wan 2.6 API introduces a re-engineered storytelling engine that generates multi-shot, 1080p videos with smooth transitions, balanced pacing, and natural camera movement. It understands storyboard-style prompts and scene descriptions, allowing developers to create connected visual narratives from text or image inputs. This makes the Wan 2.6 AI Video Generation API ideal for cinematic storytelling and short-form creative production.

Reference-Based Generation for Stable Identity and Voice via the Wan Video API

The latest Wan 2.6 API adds a powerful reference-based generation system that extracts appearance, motion style, and voice characteristics from a reference clip. The Wan Video API then applies these attributes consistently to new scenes, maintaining character and style continuity throughout the entire video. This upgrade enables character-driven storytelling and brand-consistent AI-generated videos that look realistic and professionally produced.

Extended Duration and Enhanced Temporal Stability supported by the Wan 2.6 AI Video Generation API

Wan 2.6 API extends video length to up to 15 seconds while maintaining HD clarity and frame-to-frame consistency. Improved temporal attention keeps lighting, outfit, and environmental details stable throughout motion. This longer, smoother output gives developers more flexibility to build richer narratives for commercial-grade AI video generation projects.

Integrated Audio for Realistic HD Output enabled by the AI Video Generation API

The Wan 2.6 AI Video Generation API combines native audio creation and advanced camera physics in one workflow. It produces synchronized dialogue, background music, and ambient sound with precise lip sync while executing realistic pans, zooms, and tracking shots. The result is a fully cinematic HD AI video creation experience—high-fidelity, ready-to-publish output directly from Wan 2.6 API.

How Wan 2.6 Compares to Other AI Video Models

As AI video generation continues to evolve, Wan 2.6 stands alongside several major models that each bring their own approach to duration, audio, motion, and scene control. The comparison below places Wan 2.6 next to Wan 2.5, Sora 2, Veo 3.1, and Kling 2.6, offering a clear view of their publicly available capabilities so developers and creators can understand how these systems differ and where each model fits within modern video workflows.

FeatureWan 2.6Wan 2.5Sora 2Veo 3.1Kling 2.6
Input TypesText, Image, Video ReferenceText, ImageText, ImageText, ImageText, Image
Typical Output DurationUp to ~15 seconds~8–10 secondsUp to ~25 seconds8 seconds, supports extended durations~3–10 seconds
Resolution1080p1080p1080p1080p1080p
Audio GenerationYes, native audio with lip-syncYes, supports audio generationYes, integrated audioYes, integrated audioYes, native audio and speech
Multi-Shot / Scene ControlYes, supports multi-scene promptsLimitedSupports complex scene structuresSupports multi-prompt sequencingLimited multi-scene ability
Character ConsistencyStrong multi-shot consistencyModerateStrong character continuityStable character continuityStable appearance and motion
Camera Motion ControlNatural cinematic motionLimitedAdvanced camera motionCinematic camera transitionsRealistic, physics-aware camera motion
Primary FocusCinematic narratives and reference-driven outputsCore capabilities for short videosRealistic sequences with strong coherenceControlled cinematic sequencesHigh-fidelity motion with synchronized audio

How to Use the Wan 2.6 API on Kie.ai

  • Create an Account and Retrieve Your Wan 2.6 API Key

    Begin by registering or signing in to Kie.ai, then access the API Dashboard to generate your Wan 2.6 API key. This key authenticates all requests and enables secure use of the video generation endpoints.

  • Explore the Wan Video API in the Playground

    Before integrating the API, experiment with the Playground to evaluate Wan 2.6’s behavior across text-driven, image-driven, and reference-assisted generation. This interactive environment helps you understand model performance, narrative handling, and output structure without writing any code.

  • Configure Your Environment for API Integration

    Set up your development environment using REST or the official SDK. Add your API key, define your request schema, and prepare the parameters—such as duration, input type, and output format—that align with your workflow requirements.

  • Submit Your First Wan 2.6 Generation Request

    Send a structured request specifying your prompt or media input. The API returns a task identifier that you can query until the final HD video is ready. This step validates that your integration is functioning correctly.

  • Incorporate Wan 2.6 into Your Production Workflow

    Once your setup is established, integrate the API into your application logic. Standardize prompts, refine parameters, and align generation settings with your content pipeline to support consistent and scalable video creation.
4.9/ 5
32,030 ratingsLoved by most users
How was your experience?

See What You Can Build with the Wan 2.6 Video API

Surreal Cinematic Animation with the Wan 2.6 Video API

The Wan 2.6 Video API enables creators and developers to build expressive cinematic sequences that transition smoothly across environments, perspectives, and lighting conditions. Ideal for creative storytelling and visual experimentation, the API preserves stable textures, reflective materials, and multi-shot continuity—supporting artistic short films, campaign visuals, and stylized narrative content.

Hyperreal ASMR Macro Content Using the Wan 2.6 API

For tactile, sensory-driven visuals, the Wan 2.6 API generates hyper-detailed macro scenes with precise micro-reflections, consistent depth of field, and controlled pacing. This use case fits ASMR creators, product-detail content, food close-ups, and any scenario requiring crisp, intimate textures powered by advanced AI video generation.

Commercial Product Reveal and Branding with the Wan Video API

The Wan Video API offers reliable lighting control, clean contours, and polished camera transitions—ideal for product unveilings, branded assets, and commercial motion content. From metallic surfaces to engineered objects, the system reproduces modern product aesthetics with clarity, making it a strong fit for e-commerce, marketing teams, and industrial designers.

Atmospheric Sci-Fi Worldbuilding via the Wan 2.6 Video API

Teams producing narrative, game, or film concepts can use the Wan 2.6 Video API to develop atmospheric sci-fi sequences with large-scale environments, drifting particles, and dramatic cinematic lighting. The model maintains structural coherence across both wide planetary shots and interior scenes, enabling immersive worldbuilding and high-concept storytelling.

Why Choose Kie.ai for AI Video Generation with Wan 2.6 API

Text, Image and Video Input on One Endpoint

wan/2-6-text-to-video, wan/2-6-image-to-video and wan/2-6-video-to-video run on POST /api/v1/jobs/createTask with 720p or 1080p output and Chinese or English prompts; the full parameter list is in the playground and API docs above. Pricing for each duration and resolution is shown in the pricing card above.

Output Shape Depends on the Mode

Text-to-video always outputs a 16:9 video with generated audio, and portrait output is not available. Image-to-video keeps the aspect ratio of the input image and accepts only one image per request.

Set the Resolution Explicitly

Set the resolution explicitly for text-to-video and video-to-video; image-to-video defaults to 1080p. Video-to-video is limited to 5 or 10 seconds, while the other two also allow 15.

Reference Videos, Multi-Shot and 80 Free Credits

Video-to-video takes up to 3 reference videos and generates a new video featuring their subject in the scene your prompt describes. All three models can switch between one continuous shot and multiple shots with transitions. New Kie.ai users receive 80 free credits for testing.

FAQs About Wan 2.6 API on Kie.ai

The Wan 2.6 AI Video Generation API supports text-to-video, image-to-video, and reference-driven workflows, producing up to 15 seconds of 1080p cinematic output. Multi-shot storytelling, stable character identity, coherent lighting, and native audio sync make the Wan 2.6 API suitable for creative, commercial, and narrative applications.

Yes. Kie.ai provides a free Playground where you can instantly test the Wan 2.6 Text-to-Video API, Wan 2.6 Image-to-Video API, and reference-guided generation. This lets you evaluate pacing, duration, shot transitions, and HD quality before implementing any API calls in production.

The Wan 2.6 API accepts natural language prompts, single images, and short reference videos. These inputs work seamlessly across the Wan 2.6 T2V API, Wan 2.6 I2V API, and Wan 2.6 Reference-to-Video API, allowing detailed control over motion, structure, and stylistic consistency.

The Wan 2.6 HD Video API generates 1080p video with synchronized audio, realistic camera behavior, smooth multi-shot transitions, and strong temporal stability. Outputs from Wan 2.6 T2V and I2V endpoints maintain cinematic clarity and consistent visual details across the entire sequence.

The latest Wan 2.6 API adds reference-to-video generation, 15-second extended output, enhanced temporal stability, multi-shot narrative control, improved lighting consistency, and integrated audio-visual creation. All Wan 2.6 API endpoints benefit from these upgrades, including T2V, I2V, and reference-driven workflows.

The Wan 2.6 Multi-Shot Video API is engineered to preserve facial identity, wardrobe, textures, lighting, and environmental cues across consecutive shots.

Wan 3.0 Video adds structured references, which can help maintain steadier character continuity across Wan 2.6’s multi-shot scenes. Try it at Wan 3.0 Video on Kie.ai.