wan-2.7-i2v-spicyWan 2.7 Spicy is an image-to-video model built on Alibaba's Wan 2.7 and fine-tuned for relaxed content moderation. Give it a first frame and a prompt and it returns a 2 to 15 second clip at 720p or 1080p with a generated stereo audio track — ambience, effects or voice — already inside the MP4. Pass your own audio file as audio_url and the model animates to it instead; set audio to false for a silent file. The latest upstream build made the audio fuller and the motion smoother, added anime-style and Asian-face scenes, and halved the prompt-rewriting step, so a one-line prompt is still expanded into camera, lighting and sound detail before generation (prompt_extend, on by default and free). Billing is per second of the duration you request: $0.169/s at 720p and $0.26/s at 1080p — a 5-second 1080p clip is $1.30, a 15-second one $3.90 — and failed requests are never charged. Measured on this platform: a 1080p, 2-second clip came back in 45 seconds as 1920x1080 at 30fps with AAC 44.1kHz stereo, no watermark. Where it fits: this is the premium tier of the Spicy line. Wan 2.2 Spicy is the volume engine — silent, 5 or 8 seconds, from $0.03/s, the right tool for game sprites, chat avatars, social loops and feed covers. Wan 2.7 Spicy is the clip you sell: pay-per-view and subscription content where the native audio does the immersion without a post step, branching interactive stories where every choice is one 2–15 second call, creator tools whose users publish straight from the API, and marketplaces where 1080p plus sound is the quality bar cheaper models cannot reach. Limits to design around: exactly one first frame (no last-frame interpolation — use wan-2.2-i2v-spicy for that), duration is a whole number from 2 to 15, resolution defaults to 1080p, and with prompt rewriting on the scene can drift away from the first frame within a couple of seconds — turn prompt_extend off when the exact scene must hold. We add no moderation layer of our own; the model's own safety systems still apply upstream. Complying with the law and your platform's rules, and using only subjects who have consented, is your responsibility.
JPG / PNG (no transparency) / WebP, 240–8000 px per side, max 20MB
WAV / MP3, 2–30 s, max 15MB. Leave empty and the model generates its own audio.
Generated video will appear here
Provide URLs and click Generate
Ambience, effects or voice generated with the frames as AAC stereo — or pass audio_url and the clip animates to your own sound
Whole seconds, billed on what you request: $0.169/s at 720p, $0.26/s at 1080p. A 5-second 1080p clip is $1.30
A Spicy fine-tune of Wan 2.7. We add no filter of our own; the model's own safety systems still apply upstream
Added in the latest upstream build, together with fuller audio, smoother motion and a prompt rewriter that runs in half the time
The premium tier of the Spicy line for pay-per-view clips, interactive stories and creator tools. Wan 2.2 Spicy is the silent volume engine from $0.03/s
Wan 2.7 Spicy is a Video Generation API provided by Alibaba. Wan 2.7 Spicy is an image-to-video model built on Alibaba's Wan 2.7 and fine-tuned for relaxed content moderation. Give it a first frame and a prompt and it returns a 2 to 15 second clip at 720p or 1080p with a generated stereo audio track — ambience, effects or voice — already inside the MP4. Pass your own audio file as audio_url and the model animates to it instead; set audio to false for a silent file. The latest upstream build made the audio fuller and the motion smoother, added anime-style and Asian-face scenes, and halved the prompt-rewriting step, so a one-line prompt is still expanded into camera, lighting and sound detail before generation (prompt_extend, on by default and free). Billing is per second of the duration you request: $0.169/s at 720p and $0.26/s at 1080p — a 5-second 1080p clip is $1.30, a 15-second one $3.90 — and failed requests are never charged. Measured on this platform: a 1080p, 2-second clip came back in 45 seconds as 1920x1080 at 30fps with AAC 44.1kHz stereo, no watermark. Where it fits: this is the premium tier of the Spicy line. Wan 2.2 Spicy is the volume engine — silent, 5 or 8 seconds, from $0.03/s, the right tool for game sprites, chat avatars, social loops and feed covers. Wan 2.7 Spicy is the clip you sell: pay-per-view and subscription content where the native audio does the immersion without a post step, branching interactive stories where every choice is one 2–15 second call, creator tools whose users publish straight from the API, and marketplaces where 1080p plus sound is the quality bar cheaper models cannot reach. Limits to design around: exactly one first frame (no last-frame interpolation — use wan-2.2-i2v-spicy for that), duration is a whole number from 2 to 15, resolution defaults to 1080p, and with prompt rewriting on the scene can drift away from the first frame within a couple of seconds — turn prompt_extend off when the exact scene must hold. We add no moderation layer of our own; the model's own safety systems still apply upstream. Complying with the law and your platform's rules, and using only subjects who have consented, is your responsibility. Through APIMODELS platform, you can access this model via a unified API with transparent pay-as-you-go pricing. Current pricing: 720p: $0.169, 1080p: $0.26.
Quickly generate brand promotion videos for ad campaigns and social media marketing.
Create compelling short-form video content for platforms like TikTok, Instagram, and YouTube.
Generate product feature demonstrations and tutorials to improve user conversion.
Produce course explanations, knowledge explainers, and training videos at low cost.
Wan 2.7 Spicy is available through APIMODELS at: 720p: $0.169, 1080p: $0.26. Billing is pay-as-you-go — you only pay for what you generate.
Sign up at APIMODELS, get your API key, and call our unified API endpoint. We provide detailed API documentation with code examples in cURL, Python, and Node.js.
APIMODELS offers the same Wan 2.7 Spicy model through our aggregation platform. We provide a unified API interface so you do not need separate accounts for each provider - one API key to access all models.
Per second of the duration you request: $0.169/s at 720p and $0.26/s at 1080p. A 5-second clip is $0.845 or $1.30, 10 seconds $1.69 or $2.60, 15 seconds $2.535 or $3.90. The audio track, a driving audio file, negative prompts and prompt rewriting cost nothing extra, and failed requests are not charged. Omitting resolution renders and bills at 1080p — pass 720p explicitly to pay less.
They complement each other. **Wan 2.2 Spicy** is the volume engine: $0.03/s at 480p and $0.06/s at 720p, 5 or 8 seconds only, **silent**, with first-to-last-frame interpolation — the right tool for breathing and blinking sprites in game dialogue, a library of mood clips behind a chat avatar, looping social shorts, animated feed covers, VTuber loop backgrounds and tip reactions. **Wan 2.7 Spicy** is the premium tier: 720p or 1080p, any whole number of seconds from 2 to 15, **native audio**, no last frame — for pay-per-view and subscription clips, interactive stories where each branch is one call, creator tools whose users publish straight from the API, and marketplaces where 1080p plus sound is the quality bar. Same endpoint, same key; only the model field changes.
Yes. By default the MP4 carries a stereo AAC track the model generates to match the scene (44.1kHz, 2 channels in our test). Pass audio_url — an https link to a 2–30 second WAV or MP3, at most 15MB — and the model drives the clip from your sound instead, with that audio in the output. Pass audio: false for a silent file, even if audio_url was given.
It is a community fine-tune of Alibaba's Wan 2.7 with looser content filtering than the official line; the latest upstream build also added anime-style and Asian-face scenes. We add no moderation layer of our own — prompts and reference images pass through as sent — but the model's own safety systems still apply upstream and screen both inputs and outputs, so this is not an unfiltered endpoint. Reviewing what you generate, complying with the law and your platform's rules, and using only subjects who have consented is your responsibility.
The first frame is a JPEG, PNG (no transparency), BMP or WebP with each side between 240 and 8000 pixels, an aspect ratio between 1:8 and 8:1 and at most 20MB, as a public URL or base64; exactly one frame — a last_image returns a 400. Duration is a whole number from 2 to 15; 5.5 or 16 returns a 400 rather than being rounded and billed. Measured here, a 1080p 2-second clip took 45 seconds; longer and sharper clips take longer. One more thing: with prompt rewriting on, the scene can drift away from the first frame within a couple of seconds — pass prompt_extend: false when the exact scene must hold.
On APIMODELS, Wan 2.7 Spicy runs alongside 60+ models on one API key and one balance, so choosing is about fit, not lock-in. It supports Image to Video, Native Audio, 2–15s, 720p / 1080p, Audio-driven, Relaxed Moderation, and you can weigh it on price and capability against other Video Generation models, then switch by changing a single model-name string — no new account or integration. Browse every Video Generation option with live pricing at apimodels.app/models.
Wan 2.7 Spicy supports: Image to Video, Native Audio, 2–15s, 720p / 1080p, Audio-driven, Relaxed Moderation. See the APIMODELS docs for full parameters and call examples.
Yes. APIMODELS exposes Wan 2.7 Spicy through a single unified API and one key — no separate provider accounts, and no need to handle each provider's regional network access yourself.
We support Stripe (Visa, Mastercard, and other international cards) and Alipay. Credits are available instantly after payment.