Wan 2.7
Wan 2.7 is a video generation model supporting text-to-video and image-to-video creation with audio and reference image inputs.
Wan
WaveSpeedText and image to video generation
Wan 2.7 is a video generation model published by Wan and served through the Wavespeed provider on MindStudio. It supports both text-to-video and image-to-video workflows, accepting inputs such as reference images, audio, existing video clips, and first/last frame images to guide generation. The model is tagged as a flagship release and was made available in February 2026.
Wan 2.7 is designed for creators and developers who need flexible video synthesis from multiple input types within a single model. It exposes controls for resolution, aspect ratio, duration, seed, and optional prompt expansion, giving users fine-grained control over output. Pricing ranges from $0.50 to $3.20 per generation depending on configuration, and the model operates with a context window of up to 50,000 tokens.
What Wan 2.7 supports
Text to Video
Generates video clips directly from a text prompt, with optional prompt expansion to automatically enrich short descriptions into more detailed instructions.
Image to Video
Animates a provided image into a video sequence, with support for specifying both a first frame and a last frame image to control the motion arc.
Audio Input
Accepts an audio URL as an input, enabling audio-driven or audio-synchronized video generation workflows.
Reference Image Guidance
Accepts an array of reference images to guide visual style or subject consistency across the generated video.
Video Input
Takes an array of existing video URLs as input, allowing video-to-video or continuation-style generation tasks.
Output Configuration
Exposes selectable controls for resolution, aspect ratio, and duration, plus a seed parameter for reproducible outputs.
Negative Prompting
Supports a negative prompt input to explicitly exclude unwanted visual elements or styles from the generated video.
Ready to build with Wan 2.7?
Get Started FreeCommon questions about Wan 2.7
What types of inputs does Wan 2.7 accept?
Wan 2.7 accepts text prompts, image URLs (including first and last frame images), audio URLs, arrays of reference images, and arrays of video URLs, making it suitable for a range of text-to-video and image-to-video workflows.
How much does it cost to use Wan 2.7?
Pricing for Wan 2.7 ranges from $0.50 to $3.20 per generation. The exact cost depends on the configuration options selected, such as resolution and duration.
What is the context window for Wan 2.7?
Wan 2.7 has a context window of 50,000 tokens and a maximum response size of 10,000 tokens.
When was Wan 2.7 released?
Wan 2.7 was released in February 2026 and became available on MindStudio on April 14, 2026.
Can I control the style and length of generated videos?
Yes. Wan 2.7 provides selectable parameters for resolution, aspect ratio, and duration, as well as a seed value for reproducible results and an optional prompt expansion feature to enrich shorter prompts automatically.
Documentation & links
Parameters & options
Start frame image to animate, or a reference image to guide the scene.
End frame image to define where the clip finishes.
Audio track to synchronize with the generated video.
Reference videos for character/style guidance. Refer to them as "Video 1", "Video 2", etc. in the prompt.
Reference images (max 5). Combined with videos, total must be 1-5. Referred to as "Image N" continuing after video numbering.
Describe what you want to exclude from the output.
When enabled, the model automatically enriches and optimizes your prompt before generation.
Explore similar models
WanWan 3.0
Wan 3.0 is a video generation model that converts text prompts and images into video clips with optional audio output.
WanWan 3.0 Prime
WanWan 3.0
ByteDanceSeedance 2.0 Fast Turbo
Seedance 2.0 Fast Turbo is a video generation model from ByteDance that supports text-to-video and image-to-video workflows.
GoogleVeo 3.1
Veo 3.1 is a Google video generation model that creates videos from text prompts or images, with optional audio output.
MiniMaxMiniMax H3
MiniMax H3 is a video generation model from MiniMax that supports text-to-video, image-to-video, and reference-based video creation at up to 2K resolution.
Start building with Wan 2.7
No API keys required. Create AI-powered workflows with Wan 2.7 in minutes — free.