curl-X POST 'https://pollo.ai/api/platform/v1/generation/pixverse/pixverse-v3-5/video'\-H'Content-Type: application/json'\-H'x-api-key: YOUR_API_KEY'\-d'{ "input": { "prompt": "A cinematic shot of a golden retriever running through a field of sunflowers at sunset, warm rim light, shallow depth of field", "aspectRatio": "16:9", "duration": 5, "resolution": "360p", "style": "auto", "mode": "normal", "generateAudio": true }}'
{"input":{"prompt":"A cinematic shot of a golden retriever running through a field of sunflowers at sunset, warm rim light, shallow depth of field","aspectRatio":"16:9","duration":5,"resolution":"360p","style":"auto","mode":"normal","generateAudio":true},"webhookUrl":"https://example.com/webhooks/pollo"}
模式
輸入所選端點和模式的欄位。
場地
類型
必需的
描述
prompt
string
支持
Text prompt describing the content to generate.
aspectRatio
string
不
Output aspect ratio, as width:height (e.g. 16:9).
duration
number
不
The duration of the generated video, in seconds.
resolution
string
不
Output resolution of the generated media (e.g. 720p, 1080p, 2K, 4K). Values are case-insensitive.
style
string
不
Visual style preset applied to the generation.
mode
string
不
Generation mode of the model (e.g. a faster or higher-quality tier).
seed
integer
不
Random seed for reproducible generation.
generateAudio
boolean
不
Generate natural-sounding, fitting audio for the output video.
PixVerse 3.5 supports built-in anime, 3D animation, clay, comic, and cyberpunk style presets for generating videos with different visual directions.
With PixVerse 3.5 on Pollo API, you can integrate text-to-video, image-to-video, and first-to-last frame generation, with configurable clip length, aspect ratio, and fast generation mode for each request.
Key Features of PixVerse 3.5 API
Stylized Preset Range
Support anime, 3D animation, clay, comic, and cyberpunk style presets that can be selected to define the generated video’s visual direction.
Start and End Frame Control
Supply a first image and a tail image to guide the motion and transition between the opening and ending frames.
Fast Draft Mode
Provide configurable generation modes for balancing processing speed and output resolution across preview and final-generation workflows.
Prompt and Negative Prompt Guidance
Accept prompts and negative prompts to define the scene, guide the generated content, and specify elements that should be excluded.
Flexible Framing
Support multiple aspect ratios for generating widescreen, vertical, square, and standard-format videos across different content workflows.
Use Cases of PixVerse 3.5 API
Anime Short Clips: Turn a written scene into animated anime motion for webtoon, fan-art, and character-focused creator tools using the built-in anime preset.
Generate playful clay or comic-styled sequences for education apps and social explainers that want a hand-crafted feel.
Clay and Comic Explainers:
Cyberpunk Scene Builders: Produce neon, high-contrast futuristic clips for game promos and music visuals without hand-tuning a style prompt.
Frame-to-Frame Transitions: Feed a start and end image to animate a logo morph, a before-and-after, or a controlled camera move between two known shots.
Vertical Social Drafts: Use fast mode and 9:16 framing to batch quick vertical concepts for short-form platforms before final rendering.
Still Image Animation: Bring a single illustration or photo to life as a short moving clip while keeping its composition intact.
3D Product Motion: Apply the 3D animation preset to give objects a rendered, stylized spin for storefronts and app onboarding.
How to Use PixVerse 3.5 API
Get an API Key: Create a Pollo API account and generate your API key from the developer dashboard.
Choose the Model: Send requests to the PixVerse 3.5 endpoint at /generation/pixverse/pixverse-v3-5.
Add Your Inputs: Provide a prompt for text-to-video, or an image URL (plus optional imageTail) for image-to-video, then set style, resolution, length (5 or 8), and mode.
Generate and Retrieve: Submit the task, poll the returned task status, and download the finished video once it succeeds.
Prompting Best Practices for PixVerse 3.5 API
Select a style preset first, then describe the subject, action, and framing clearly. Use the preset to define the visual direction and the negative prompt to specify elements that should be excluded.
A Simple Prompt Formula
Style preset + subject + specific action + camera or framing + setting and mood
What You Should Notice
Lead With A Preset: Choose anime, clay, comic, cyberpunk, or 3D at the beginning of the request to guide the clip’s visual direction.
Keep Actions Simple: Describe one primary movement per clip to provide clearer motion guidance across 5- or 8-second videos.
Use Negative Prompts: Specify unwanted elements, such as text overlays or extra limbs, to guide what should be excluded from the output.
Match Mode To Goal: Use fast mode for prompt and composition previews, and normal mode when higher-resolution output is required.
Set Framing Early: Choose an aspect ratio that matches the target format to guide the composition of vertical, square, or landscape videos.
Example Prompts
Anime Action Beat
"Anime style: a young courier sprints across rooftops at dusk, cloak snapping in the wind, then leaps to the next ledge. Low tracking camera, warm orange sky, vertical 9:16 framing."
Claymation Kitchen Scene
"Clay style: a round little chef stirs a bubbling pot, steam curling upward, then lifts the spoon for a proud taste. Soft studio light, gentle stop-motion feel, cozy handmade textures."
Cyberpunk Street Loop
"Cyberpunk style: a lone rider glides a hoverbike down a rain-slick alley lined with flickering neon signs, reflections rippling on wet pavement. Slow forward dolly, deep blue and magenta glow."
Frame Transition Reveal
"3D animation style: a closed metallic capsule slowly unfolds, panels rotating outward to reveal a glowing core inside. Centered camera, clean gradient backdrop, smooth mechanical motion between start and end frames."
PixVerse 3.5 vs Pika 2.2 vs Kling 2.1
Capability
PixVerse 3.5
Pika 2.2
Kling 2.1
Built-in style presets
✅ Anime, 3D, clay, comic, cyberpunk
✅ Limited stylized options
❌ Prompt-driven only
Start and end frame control
✅ First and tail image
✅
✅
Text and image inputs
✅
✅
✅
Resolution options
360p to 1080p
Up to 1080p
Up to 1080p
Clip length
5 or 8 seconds
Short clips
5 or 10 seconds
Fast draft mode
✅
❌
❌
Recommended For
Stylized short-form and quick drafts
Playful effect-driven clips
Realistic character motion
Why Choose PixVerse 3.5 API?
PixVerse 3.5 is suitable for applications that require built-in style presets, fast generation modes, and transitions between starting and ending frames.
Support for anime, clay, comic, 3D animation, and cyberpunk styles makes it suitable for stylized animation, social content, character videos, and transition-based workflows.
Through Pollo API, PixVerse 3.5 can be accessed alongside other video and image models through a single API, simplifying model evaluation, integration, and deployment.
PixVerse 3.5 API FAQs
What is PixVerse 3.5?
PixVerse 3.5 is an upgraded PixVerse video model with improved quality and more customization, supporting text-to-video, image-to-video, and start-to-end frame generation across five style presets.
Does PixVerse 3.5 API support image inputs?
Yes. You can animate a single image, and add a tail image so the model generates a controlled transition between your chosen start and end frames.
What styles can PixVerse 3.5 produce?
It offers auto, anime, 3D animation, clay, comic, and cyberpunk presets, so a stylized look is selected directly instead of being described in the prompt.
What resolutions and lengths are available?
Clips run 5 or 8 seconds at resolutions from 360p up to 1080p. Note that 1080p is not available in fast mode or for 8-second clips.
What is fast mode used for?
Fast mode speeds up generation for quick previews and iteration, trading the highest resolution for turnaround, which suits drafting before a final render.
Why run PixVerse 3.5 through Pollo API?
Pollo API provides access to PixVerse 3.5 alongside other video and image models through a single integration, with API documentation and task status tracking for asynchronous video generation.
