curl-X POST 'https://pollo.ai/api/platform/v1/generation/minimax/hailuo-2-3/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", "image": "https://example.com/first-frame.jpg", "resolution": "768p", "duration": 6 }}'
Optionally include webhookUrl to receive a callback when the task succeeds or fails.
Optional request field
{"input":{"prompt":"A cinematic shot of a golden retriever running through a field of sunflowers at sunset, warm rim light, shallow depth of field","image":"https://example.com/first-frame.jpg","resolution":"768p","duration":6},"webhookUrl":"https://example.com/webhooks/pollo"}
Schema
Input fields for the selected endpoint and mode.
Field
Type
Required
Description
prompt
string
No
Text prompt describing the content to generate.
image
string
Yes
First-frame / reference image URL (HTTP(S).
resolution
string
No
Output resolution of the generated media (e.g. 720p, 1080p, 2K, 4K). Values are case-insensitive.
Hailuo 2.3 turns a written brief into motion that follows real physical forces. Momentum, balance, and contact between objects remain believable on repeated viewing.
Running Hailuo 2.3 on Pollo API lets developers wire text-to-video and image-to-video into their products, picking 768P or 1080P and 6 or 10 second clips per call.
Key Features of Hailuo 2.3 API
Force-Aware Motion
Render weight transfers, object reactions to impact, and limb follow-through under gravity, so leaps, spills, and swings resolve as viewers instinctively expect.
Emotion in the Face
Align subtle brow tension, gaze shifts, and mouth shape with the described feeling, keeping close-ups readable and speaker-driven clips convincing without extra rerolls.
Prompt Fidelity
Follow descriptions of subjects, actions, and framing closely, reducing iteration cycles and making outputs predictable within automated workflows.
Broad Art Direction
Maintain a chosen look, including crisp realism, anime, or a distinctive illustrated style, to support varied creative briefs through one integration.
Frame-Steady Clips
Keep both 6- and 10-second outputs coherent shot to shot, with stable subjects and controlled camera motion for social, explainer, and preview formats.
Use Cases of Hailuo 2.3 API
Physical Comedy Beats: Generate clips where timing and gravity carry the joke, like a stack of books toppling or a chair tipping back and catching.
Dialogue Reaction Faces: Render close-ups that move from confusion to realization, feeding chat apps, dubbing tools, and reaction-based content engines.
Anime Moment Generators: Convert prompts into stylized animated beats for manga, webtoon, and fan-art tools that need a consistent drawn look.
Instrument and Craft Motion: Show hands playing keys, kneading dough, or throwing clay with movement that reads as physically grounded for tutorial products.
Still-to-Motion Animation: Send a single image and set it moving into a short shot while the original framing and subject stay intact.
Storyboard Previews: Let creative teams see a described shot in motion fast, trusting fidelity to the brief before committing crew or budget.
Nature and Weather Clips: Produce falling leaves, rippling water, or drifting snow where believable secondary motion matters more than staged spectacle.
How to Use Hailuo 2.3 API
Get an API Key: Sign up for Pollo API and create your API key from the developer dashboard.
Choose the Model: Point your request at the Hailuo 2.3 endpoint /generation/minimax/minimax-hailuo-2.3.
Add Your Inputs: Pass a prompt for text-to-video or an image URL for image-to-video, then set resolution (768P or 1080P) and length (6 or 10).
Generate and Retrieve: Submit the task, poll the returned status, and pull the finished video once it reports success.
Prompting Best Practices for Hailuo 2.3 API
Treat the prompt as a physics note. Spell out what moves, the force driving it, and the reaction you expect, then add the camera and mood so the model can ground the shot.
A Simple Prompt Formula
Subject + action with a clear force or intent + reaction or facial cue + camera behavior + setting, style, and lighting
What You Should Notice
Name the Physics: Say how something moves, such as topples, rebounds, or slows, so the model has a concrete motion target.
Cue the Expression: Call out the gaze or feeling when a close-up rests on a character's face.
Lock the Style First: State realism, anime, or a specific illustration style early to keep the look steady across frames.
Stay on One Action: Build each clip around a single movement so the 6 or 10 seconds hold together.
Fit Resolution to Intent: Reach for 1080P on detail-heavy shots and 768P for quick drafts and cheaper iteration.
Lean on Prompt Optimization: The built-in optimizer refines terse prompts, so short briefs still resolve with sensible staging.
Example Prompts
Grounded Physical Gag
"A tall stack of paper cups on a café counter wobbles as someone bumps the table, sways twice, then collapses across the surface in a slow slide. Static camera, warm morning light, natural realistic style."
Character Realization
"A student reading a letter frowns, eyes scanning line by line, then lips part and shoulders drop as the news sinks in. Tight close-up, soft window light, slow push-in, grounded realistic style."
Anime Action Beat
"A young archer draws a glowing bowstring on a cliff edge, wind sweeping cloak and hair, then releases as a streak of light arcs outward. Wide shot, vivid anime style, dramatic backlight, camera easing forward."
Craft in Motion
"Hands center wet clay on a spinning wheel, fingers press inward and the walls rise evenly into a bowl, water glinting on the surface. Overhead angle, soft studio light, clean realistic style focused on believable motion."
Hailuo 2.3 vs Kling 2.6 vs Seedance 1.5 Pro
Capability
Hailuo 2.3
Kling 2.6
Seedance 1.5 Pro
Force and collision realism
✅ Strong weight and impact response
✅ Fluid motion, lighter physics focus
✅ Smooth but stylized motion
Facial expressiveness
✅
✅
✅
Realism plus anime range
✅
✅
✅
Resolution options
768P and 1080P
Up to 1080P
Up to 1080P
Clip length
6 or 10 seconds
Short directed clips
Short multi-shot clips
Recommended For
Physics-driven action and reactions
Character motion and dialogue
Coherent multi-shot scenes
Why Choose Hailuo 2.3 API?
Hailuo 2.3 suits products where movement has to look physically honest and faces have to carry feeling, spanning grounded action, quiet reactions, and stylized anime output.
On Pollo API, one key unlocks Hailuo 2.3 next to 300+ leading video and image models, with clean docs, task status polling, webhooks, and generation.
Hailuo 2.3 API FAQs
What is Hailuo 2.3?
Hailuo 2.3 is a Minimax video generation model built around physical accuracy, expressive faces, and prompt-faithful motion, supporting text-to-video and image-to-video across realistic and anime styles.
Does Hailuo 2.3 API support text and image inputs?
Yes. Generate from a text prompt or animate an image URL, which covers both idea-first and asset-first workflows from a single endpoint.
What resolutions and clip lengths does it offer?
Hailuo 2.3 outputs 768P or 1080P, with clip lengths of 6 or 10 seconds chosen per request.
How good is its motion handling?
Its core strength is physical plausibility, so weight, momentum, and impact read believably, which keeps action shots and object motion from feeling weightless.
Can it produce anime-style video?
Yes. Beyond realism, Hailuo 2.3 holds anime and other distinctive art styles consistently, making it a fit for tools serving stylized content communities.
Why run Hailuo 2.3 through Pollo API?
Pollo API gives you one integration for Hailuo 2.3 and 300+ other models, with clear documentation, task tracking, webhooks, and lower generation costs than comparable providers.
