Core capability
Three Wan routes, three different filter situations
Production window: the 21 days to 9 September 2026 unless a 60-day window is named. Rates are rounded down. Wan 2.2 Spicy image-to-video: 16,059 jobs, 658 creators, 98% success, zero content-safety stops. Wan 2.7 Spicy image-to-video (60 days): 374 jobs, 95 creators, 86% success, zero content-safety stops. Wan 3.0 reference-to-video: 6,901 jobs, 349 creators, 94% success, 155 review stops (2.2%). So the honest ranking for “will it render what I asked” is 2.2 Spicy, then 2.7 Spicy, then 3.0; the ranking for length, audio and control runs the other way.
What Wan 3.0 adds over the Spicy routes
Alibaba Cloud documents Wan 3.0 as a single reference-based model covering text-to-video, first-frame and first-plus-last-frame image-to-video, and reference-to-video with up to 10 images, 5 video clips and 5 audio clips. Output runs to 30 seconds at 30 fps with native audio. The Spicy routes are image-to-video only: Wan 2.7 Spicy offers 720p or 1080p at 2, 5, 10 or 15 seconds; Wan 2.2 Spicy is the short-loop workhorse. If your brief needs a second character, dialogue or a scene longer than 15 seconds, only Wan 3.0 can do it.
Where the Wan 3.0 review actually bites
Of the 155 stopped Wan 3.0 reference jobs, 126 were rejections of the rendered video, not of the prompt. The pattern in the logs is consistent: fast, multi-position sequences with faces turned away or anatomy errors get flagged; slow, continuous, face-visible scenes pass. That is why the block rate is 2.2% and not 20%. Treat the review as a quality gate you can prompt around, and treat a second rejection of the same brief as a signal to move that brief to a Spicy route.
Versions and current access
All three Wan routes named here are available in Makeartly Studio now; the CTA preselects Wan 3.0 when the live catalogue confirms the model ID.
Evidence
What “Wan 3.0 Spicy” means on other sites
Hosts that advertise a Wan 3.0 Spicy or Wan 3 Uncensored tier are usually running the open weights themselves with the prompt filter disabled, sometimes at 720p and without audio. That is a legitimate product, but read the fine print on duration, resolution and whether the reference modes exist. Makeartly’s Wan 3.0 route keeps the official 1080P tier, adds 2K and 4K upscale tiers, keeps audio and exposes reference-to-video; the trade is that one upstream output review remains in the loop.
Deployment filtering and adult-content boundaries
Makeartly does not add a separate mainstream adult-content refusal layer for verified adult creators; this deployment-level statement does not mean the upstream official model is uncensored or guarantee prompt acceptance. Adult workflows are for adults only. CSAM and all under-18 sexual content are prohibited. Consent and likeness authorization are required, and non-consensual intimate content, unauthorized real-person deepfakes, and sexual impersonation are prohibited.
A simple routing rule for adult work
Explicit, short, single-character loop from one still: Wan 2.2 Spicy, about 44 seconds median render and the lowest quote of the three. Explicit, single-character, up to 15 seconds at 1080p: Wan 2.7 Spicy. Multi-reference scene, dialogue or breathing audio, or anything over 15 seconds: Wan 3.0, starting at 720P for 5 seconds. Creators who follow that split spend the least on refunded runs, because each route is used for the thing it does not refuse.
Use cases
You want the most explicit result with the least friction
Choose Wan 2.2 Spicy image-to-video: 16,059 runs in three weeks and not one content-safety stop. Accept that clips are short and silent.
You want a 20-second scene with two characters and audio
Choose Wan 3.0 reference-to-video, keep the action continuous and faces visible, and expect roughly one review stop in 45 runs.
- Decide the brief first: one still and a short loop, or a longer multi-reference scene.
- For the short explicit loop open Wan 2.2 Spicy or Wan 2.7 Spicy image-to-video and upload one fictional adult still.
- For the longer scene open Wan 3.0, upload references with the face visible and write a motion-first prompt.
- Run 5 seconds at 720P first and read the credit quote before submitting.
- If Wan 3.0 blocks the rendered result twice, move that exact brief to a Spicy route rather than rewording it.
Limitations and decision checks
Acceptance test each route on your own brief
Run the same 5-second brief on Wan 2.2 Spicy and on Wan 3.0 as an acceptance test, compare motion quality and whether either result was refused, stop using the route that blocked it, and choose the other for the rest of that project.
The production rates above are averages across hundreds of creators; your brief can sit anywhere inside them. The Spicy routes are image-to-video only: Neither Wan 2.2 Spicy nor Wan 2.7 Spicy accepts text-only prompts, multiple references or audio input. If the brief needs those, the comparison is Wan 3.0 against Seedance or Kling, not against the Spicy routes.
