# Wan 2.7 > Wan 2.7 is Alibaba's AI video model, released in April 2026 as a full upgrade of Wan 2.6, that makes 2–15 second clips from text, a start and end frame, or up to 5 reference images and clips, and can edit and extend existing videos. On Kenerate AI it costs 30 credits a second at 720p and 45 at 1080p — 150 credits for a 5-second 720p clip — and the Wan 2.7 and Wan 2.7 Pro image models make stills from 9 and 22.5 credits. URL: https://kenerateai.com/models/alibaba/wan-2-7 Publisher: Kenerate AI (https://kenerateai.com) Last updated: 18 October 2026 ## Key facts - Official name: Wan 2.7 (Wan2.7) - API ID: alibaba/wan-2.7 - Maker: Alibaba (Tongyi Wanxiang) - Released: April 2026 - Clip length: 2–15 s · reference 2–10 s - Resolution: 720p · 1080p - Aspect ratios: 16:9 · 9:16 · 1:1 · 4:3 · 3:4 - Frames: Start frame + optional end frame - References: Up to 5 images and clips - On Kenerate: 30 · 45 credits a second ## Steps 1. Open the video studio — Go to /app/video and pick Wan 2.7 (the button on this page selects it). Video needs any one-time credit pack. 2. Choose a mode — Text to video, image to video (add a second image as the end frame), reference to video for up to 5 images and clips, video edit, or extend. 3. Describe the motion — For image to video, describe the movement and the camera, not what's already in the frame. Attach a wav or mp3 if lips or timing should follow it. 4. Pick length and resolution — 2–15 seconds (2–10 for references), 720p to test and 1080p to finish. The credit cost shows before you generate. ## In depth ### What each Wan 2.7 clip costs Wan 2.7's price on Kenerate depends on mode, length and resolution: - 3 s at 720p — 90 credits. A short b-roll cut or a quick motion test. - 5 s at 720p — 150 credits; 15 s at 1080p — 675 credits for the final render. - Reference to video, 5 s — 300 credits at 720p, 480 at 1080p. - Video edit of a 5 s clip — 300 credits at 720p, 450 at 1080p (input and output seconds). - Extend — the final length you ask for, 5–15 s, at 30 or 45 credits a second. ### What changed from Wan 2.6 Wan 2.6 (December 2025) brought 15-second clips and reference to video. Wan 2.7 (April 2026) keeps both and adds first-and-last-frame image to video, audio-driven generation, video continuation and a video edit mode that changes a clip from a prompt, a reference image or a style [1]. API platforms list it at 2–15 seconds, 720p or 1080p, in five aspect ratios [2], at the same per-second rate as 2.6 [3]. On Kenerate the practical difference is control: you can fix the last frame, cut clips as short as 2 seconds and work in square or 4:3, all at Wan 2.6's price. ### Prompting Wan 2.7 For image to video, describe motion plus camera — "she spins and throws the sleeves into the air, slow push-in" — and leave out what the frame already shows. With an end frame, describe the change between the two images, not either image. For text to video, write subject, scene and action, then light and lens, and for multi-shot clips list the shots in order with the same character description each time. References: [1] Wan2.7 is now available in ComfyUI via Partner Nodes — Comfy blog (3 Apr 2026): https://blog.comfy.org/p/wan27-is-now-available-in-comfyui [2] Wan 2.7 model docs — Prodia: https://docs.prodia.com/models/wan-2-7/ [3] alibaba/wan-2.7 image-to-video — WaveSpeed: https://wavespeed.ai/models/alibaba/wan-2.7/image-to-video ## FAQ Q: What is Wan 2.7? A: Wan 2.7 is Alibaba's AI video model from April 2026, a comprehensive upgrade of Wan 2.6. It does text to video with multi-shot narration, image to video from a first frame or a first and last frame, audio-driven video, reference to video that keeps a subject's look across scenes, video continuation and video editing. Alibaba also released Wan 2.7 image models alongside it. Q: Is Wan 2.7 open source? A: Not as far as Alibaba has published: only Wan 2.1 and Wan 2.2 have open weights under Apache 2.0. Wan 2.7 is offered as a hosted model through Alibaba Cloud and API partners — even ComfyUI runs it through Partner Nodes, which call an API rather than local weights. You run it online, for example in Kenerate's video studio. Q: How much does Wan 2.7 cost on Kenerate? A: It's billed per second: 30 credits at 720p and 45 at 1080p, so a 5-second clip is 150 or 225 credits and 15 seconds at 1080p is 675. Reference to video costs 300 credits for 5 seconds at 720p (480 at 1080p). Video edit bills the input and output seconds. Credit packs are one-time and start at $15 for 1,300 credits. Q: Does Wan 2.7 support a first and last frame? A: Yes. In image to video, upload a start image and a second image as the end frame, and Wan 2.7 generates the motion between them. It's the best way to land a product reveal, a before-and-after or a transition into the next shot. Wan 2.6 doesn't take an end frame. Q: How long can Wan 2.7 videos be? A: Up to 15 seconds per clip, in any whole number of seconds from 2. Reference to video runs 2–10 seconds. Video extend lets you set a final length of 5–15 seconds from an existing clip, and for longer edits you can start a new clip from the last frame of the previous one. Q: Does Wan 2.7 generate sound? A: Sources differ, and on Kenerate it has no sound switch. What it reliably does is follow an audio track: attach a wav or mp3 in image to video or extend and the motion and lip-sync are timed to it. If you want effects, ambience or speech generated from the prompt alone, Wan 3.0 is the more predictable choice. Q: Can Wan 2.7 edit a video? A: Yes. Video edit takes a 2–10 second clip plus a text prompt and up to 3 reference images, and returns a changed or restyled version the same length as the input. Because input and output are both billed, a 5-second edit costs 300 credits at 720p or 450 at 1080p. Q: What is the Wan 2.7 image model? A: It's Alibaba's image model from the same family, listed in Kenerate's image studio as Wan 2.7 and Wan 2.7 Pro. Both have a thinking mode for better-planned compositions and take up to 9 reference images for edits. Wan 2.7 costs 9 credits an image at sizes up to 2048 px; Wan 2.7 Pro costs 22.5 and goes up to 4K. Q: Is Wan 2.7 free? A: Not on Kenerate's free plan: video needs a paid plan, which means buying any one-time credit pack. New accounts get starter credits for images on free-plan models. The Wan 2.7 image models are not free-plan models either. Q: Wan 2.7 vs Wan 2.6 vs Wan 3.0 — which should I use? A: Wan 2.7 costs the same as Wan 2.6 per second and adds end frames, video editing, 2-second minimum lengths and square and 4:3 shapes, so it's the better pick for most image-to-video work. Wan 3.0 is the choice for clips up to 30 seconds with sound generated from the prompt, and Wan 2.5 for the cheapest 480p drafts. ## Related pages - Wan AI video generator: https://kenerateai.com/wan-ai-video-generator - Wan 2.6: https://kenerateai.com/models/alibaba/wan-2-6 - Wan 3.0: https://kenerateai.com/wan-3-0 - Wan image models: https://kenerateai.com/model/wan - All AI video models: https://kenerateai.com/ai-video-models ## Sources - Wan2.7 is now available in ComfyUI via Partner Nodes — Comfy blog (3 Apr 2026): https://blog.comfy.org/p/wan27-is-now-available-in-comfyui - Wan 2.7 model docs — Prodia: https://docs.prodia.com/models/wan-2-7/ - alibaba/wan-2.7 image-to-video — WaveSpeed: https://wavespeed.ai/models/alibaba/wan-2.7/image-to-video