Wan AI 2.7 Image-to-Video: Image to Video Online
Bottom line
Image-to-video with first/last-frame control, 5-15s shots at 720p/1080p, from roughly 6.4 credits/second. Upload a starting frame, choose Wan AI 2.7 Image-to-Video in the Image to Video tool, and animate it on VdoBloom — the credit price appears before every generation.
Wan 2.7 Image-to-Video supports first- and last-frame input: give it an opening image alone and it animates forward freely, or supply both endpoints and the model builds the motion that connects them. That second mode turns image-to-video from a lucky draw into a shot you actually planned -- useful for transitions, transformations, and any move that has to land on a specific composition.
Clips come in 5, 10 or 15-second lengths at 720p or 1080p, giving longer motions room to resolve instead of cutting off mid-arc. Pricing runs roughly 6.4 credits/second at 720p and 9.5-9.6 credits/second at 1080p -- notably cheaper than the older Wan 2.5 Image-to-Video (13-23 credits/second) for a comparable spec, and slightly less than Wan 2.6 Image-to-Video (roughly 7-8 to 12-12.7 credits/second), which also lacks the endpoint-frame control.
The HOT content label applies, meaning filtering sits at the permissive end and prompts refused elsewhere often run here. Alibaba has not published Wan 2.7's weights, so VdoBloom's hosted access is the way to run it without a local setup. As with every model on the platform, the credit cost for your selected duration and resolution is shown before generation, so keyframe experiments are priced before they start.
What it costs on VdoBloom
Wan 2.7 Image-to-Video bills per second: roughly 6.4 credits/second at 720p, 9.5-9.6 credits/second at 1080p. Worked examples: 5s/720p = 32 credits, 10s/720p = 64 credits, 15s/720p = 95 credits; 5s/1080p = 48 credits, 10s/1080p = 95 credits, 15s/1080p = 143 credits.
Wan 2.7 Image-to-Video vs Wan 2.5 and Wan 2.6
All three animate a single uploaded image, but price and features diverge. Wan 2.5: 5 or 10 seconds, 720p/1080p, 13-23 credits/second, no endpoint control. Wan 2.6: 5/10/15 seconds, 720p/1080p, roughly 7-8 to 12-12.7 credits/second, no endpoint control. Wan 2.7: same 5/10/15-second range, roughly 6.4-9.6 credits/second (the cheapest of the three), plus first- and last-frame control. Unless you're locked into an older pipeline, Wan 2.7 is the strongest default for new image-to-video work on VdoBloom.
What creators use Wan AI 2.7 Image-to-Video for
- Build a transformation shot by supplying a before image as the first frame and an after image as the last.
- Animate a key visual into a 15-second 1080p clip that ends on a planned composition.
- Create seamless transitions between two storyboard frames for a longer edit.
- Animate from a single starting image when you don't need endpoint control, at a lower cost than Wan 2.5 or 2.6.
How to use Wan AI 2.7 Image-to-Video on VdoBloom
- 1Open the Image to Video tool and select Wan AI 2.7 Image-to-Video from the model picker.
- 2Upload your starting image and describe the motion you want in the prompt.
- 3Choose your duration (5–15s), quality (720p/1080p), review the credit cost, and generate.
Frequently asked questions
How much does Wan 2.7 Image-to-Video cost?
Roughly 6.4 credits/second at 720p, 9.5-9.6 credits/second at 1080p. A 5-second 720p clip is 32 credits, 10 seconds is 64, 15 seconds is 95; at 1080p those are 48, 95 and 143 credits.
How does first- and last-frame control work?
You provide a starting image, and can optionally add an ending frame too. With both endpoints set, the model generates the motion connecting them, which makes transitions and transformations far more predictable than open-ended animation from a single image.
Is Wan 2.7 open source, or can I run it locally?
No -- Alibaba hasn't released Wan 2.7's weights the way it did for Wan 2.1/2.2, so there's no local install path. VdoBloom's hosted access, billed in credits, is the way to use it.
What clip lengths and resolutions does it support?
5, 10 or 15 seconds at 720p or 1080p. Longer durations give keyframed moves time to resolve naturally, and the credit cost for each combination is shown before you render.
What does the HOT content label mean for this model?
HOT is VdoBloom's most permissive filtering tier, so uploaded frames and motion prompts face lighter moderation than on STRICT-rated models. Platform-wide rules on illegal content still apply.
Should I use Wan 2.7 or Wan 2.6 for image-to-video?
Wan 2.7 is the newer generation, costs less per second (roughly 6.4 vs 7-8 credits/second at 720p), and adds first- and last-frame control. Wan 2.6 offers the same 5-15 second durations and 720p/1080p tiers without endpoint control. There's no reason to prefer 2.6 unless a project is already built around it.
How much cheaper is Wan 2.7 than Wan 2.5 Image-to-Video?
Meaningfully: Wan 2.5 costs 13-23 credits/second depending on resolution, versus Wan 2.7's roughly 6.4-9.6 credits/second -- about half, at the same 720p/1080p tiers, plus first/last-frame control that 2.5 doesn't have. A 5-second 720p clip is 65 credits on 2.5 versus 32 on 2.7.
Does Wan 2.7 Image-to-Video generate audio?
The catalog description for this variant doesn't list native audio, unlike Wan 2.7 Text-to-Video and Wan 2.6 Flash. If sound matters, check the generation settings in the composer before running a longer clip.
What's the difference between Wan 2.7 Image-to-Video and Wan 2.7 Reference-to-Video?
Image-to-Video treats your upload as the literal first (and optionally last) frame of the clip. Reference-to-Video instead uses one or more reference images or a reference video to guide a subject's identity and style while your prompt drives the action -- better for keeping a character consistent across many separate generations rather than animating one specific shot.
How do I plan a transformation shot with first/last frame?
Upload the starting state as the first frame and the end state as the last frame, then write the prompt to describe the transition itself (the how) rather than either end state (the what) -- the images already establish those. This is the mode to reach for when a move has to land on a specific composition rather than drift freely.