Wan 2.7

Enhanced Video Generation Quality

Improves fundamental output aspects including visual sharpness, motion smoothness, style adherence, audio synchronization, and subject continuity for higher production value.

Structured Generation Control

Provides precise clip definition through first-frame and last-frame inputs, plus 9-grid image sequences, enabling more controllable video generation workflows.

Multi-Reference Input Support

Anchors output using real-person images, up to five video references, and voice inputs to maintain consistent appearance, motion style, and audio direction.

Instruction-Based Video Editing

Modifies existing video content via natural language commands to alter elements like backgrounds, lighting, or styling without regenerating the entire clip.

Video Recreation and Remixing

Preserves the original motion structure while changing style, subject, or context, supporting iterative versioning and creative adaptation of existing footage.

• Short-form narrative creators: Use first-frame and last-frame generation for precise scene arcs with stable character continuity.
• Marketing teams: Produce 1080P social media ads with synchronized audio in 2-15 second durations.
• Character IP developers: Maintain appearance and voice consistency across videos using subject plus voice reference.
• Video editors: Modify existing video through natural language instructions to change backgrounds or lighting.
• Multi-reference producers: Combine real-person images and up to 5 video references for controlled style and motion.

When is Wan 2.7 expected to launch?

Wan 2.7 is planned for launch within March 2026. This scheduled release indicates the model is currently in development and not yet publicly available. Users can monitor the official Wan2.7 website for updates on availability and access options as the launch approaches.

How can users access Wan 2.7 for free?

New users receive 30 free credits upon signing up on the Wan 2.7 website. These credits allow for initial experimentation with video generation and editing features. The free tier provides a way to evaluate the wan video generator's capabilities without financial commitment before considering paid plans.

What makes Wan 2.7 different from previous Wan versions?

Wan 2.7 is a major upgrade over Wan 2.6, improving visual quality, audio fidelity, motion dynamics, stylization control, and subject consistency. New features include first-frame and last-frame generation, 9-grid image-to-video, subject plus voice reference, instruction-based editing, and video recreation. It also supports up to 5 video references, real-person image input, 2-15 second duration, and 1080p resolution.

What is the maximum video duration for Wan 2.7?

Wan 2.7 supports video generation with durations from 2 to 15 seconds. This maximum length is tailored for short-form content such as social media clips and advertisements, ensuring high-quality 1080p output within a concise timeframe suitable for modern digital platforms and marketing workflows.

What resolutions does Wan 2.7 support?

Wan 2.7 generates videos at 1080p resolution, providing high-definition output. This resolution standard ensures clarity and detail for professional applications, including marketing materials, narrative shorts, and social media content, while maintaining compatibility with common display formats and streaming platforms.

What kinds of inputs and references does Wan 2.7 support?

Wan 2.7 accepts various inputs for video generation: text prompts for text-to-video, single images or 9-grid image layouts for image-to-video, and up to five video references for reference-to-video workflows. It also supports real-person image input and subject plus voice reference to anchor appearance and vocal direction across generated clips.

How does Wan 2.7 handle audio synchronization for generated videos?

Wan 2.7 incorporates native audio synchronization, generating speech that matches lip movements naturally. This built-in feature eliminates the need for separate audio post-production, as the model aligns dialogue with visual elements during generation, enhancing realism for character-led and voice-consistent video content.

What features support character consistency in Wan 2.7?

Character consistency in Wan 2.7 is supported through subject plus voice reference, which maintains both appearance and vocal direction across clips. Additionally, multi-video reference input (up to five references) and 9-grid image-to-video provide structured control over subject identity, motion, and scene style, ensuring uniform character representation.

Can Wan 2.7 edit existing video?

Yes, Wan 2.7 can edit existing videos using instruction-based editing through natural language commands. Users can describe changes such as background replacement or lighting adjustments, and the model applies edits while preserving motion structure. The video recreation feature also allows remixing by switching subjects or styles without full regeneration.

Does Wan 2.7 provide an API for developers?

Wan 2.7 offers API access, enabling developers to integrate its video generation and editing capabilities into custom applications and automated workflows. API documentation and endpoints are available on the official website, facilitating technical implementation for scalable video production solutions and enterprise use cases.

Wan 2.7 is an AI video generator that produces short-form videos with enhanced visual quality, audio synchronization, and motion control. It supports text-to-video, image-to-video, and reference-to-video workflows for professional content creation.

• Access the Wan 2.7 platform and create a free account to receive 30 initial credits, allowing immediate video generation without upfront cost.
• Select the desired generation mode: text-to-video for prompt-driven scenes, image-to-video for single image animation, or reference-to-video for multi-video guidance.
• Enter a descriptive text prompt or upload source images/videos. For reference modes, add up to five video inputs to define style, motion, and subject appearance.
• Configure output settings, including duration (2-15 seconds), resolution (1080p), and optional controls like first-frame/last-frame or 9-grid image inputs for precise scaffolding.
• Start the generation process. Wan 2.7's model processes inputs, typically completing a 15-second 1080p clip in approximately 95 seconds under standard settings.
• Evaluate the generated video for visual sharpness, audio sync, motion fluidity, and subject consistency across frames, noting areas for potential refinement.
• Apply instruction-based editing by submitting natural language commands to alter specific elements, such as changing backgrounds or lighting, without regenerating the entire video.
• Export the final video in 1080p format. Utilize video recreation to preserve motion while altering style or subject, enabling efficient versioning for different contexts.