Mochi-1

mochi-1

Mochi 1 preview is an open state-of-the-art video generation model with high-fidelity motion and strong prompt adherence in a preliminary evaluation.

H100 80GB

Fast Inference

REST API

Try in Console API Docs Examples

Model Information

Response Time~261 sec

StatusActive

Version

0.0.1

Updated4 days ago

Live Demo

Average runtime: ~261 seconds

Input

Configure model parameters

Prompt

Focus on a single, central subject. Structure the prompt from coarse to fine details. Start with 'a close shot' or 'a medium shot' if applicable. Append 'high resolution 4k' to reduce warping

Close-up of a chameleon's eye, with its scaly skin changing color. Ultra high resolution 4k.

Guidance Scale

The guidance scale for the model

5.5

Seed

Random seed

Enter your seed

Output

View generated results

Result

Preview, share or download your results with a single click.

Cost is calculated based on execution time.The model is charged at $0.0018 per second. With a $1 budget, you can run this model approximately 2 times, assuming an average execution time of 261 seconds per run.

API Reference

View Full Documentation

Prerequisites

Create an API Key from the Eachlabs Console
Install the required dependencies for your chosen language (e.g., requests for Python)

API Integration Steps

1. Create a Prediction

Send a POST request to create a new prediction. This will return a prediction ID that you'll use to check the result. The request should include your model inputs and API key.

import requests
import time

API_KEY = "YOUR_API_KEY"  # Replace with your API key
HEADERS = {
    "X-API-Key": API_KEY,
    "Content-Type": "application/json"
}

def create_prediction():
    response = requests.post(
        "https://api.eachlabs.ai/v1/prediction/",
        headers=HEADERS,
        json={
            "model": "mochi-1",
            "version": "0.0.1",
            "input": {
  "fps": "30",
  "seed": null,
  "prompt": "Close-up of a chameleon's eye, with its scaly skin changing color. Ultra high resolution 4k.",
  "num_frames": "163",
  "guidance_scale": "6",
  "image_prompt_strength": "0.1"
}
        }
    )
    prediction = response.json()
    
    if prediction["status"] != "success":
        raise Exception(f"Prediction failed: {prediction}")
    
    return prediction["predictionID"]

2. Get Prediction Result

Poll the prediction endpoint with the prediction ID until the result is ready. The API uses long-polling, so you'll need to repeatedly check until you receive a success status.

def get_prediction(prediction_id):
    while True:
        result = requests.get(
            f"https://api.eachlabs.ai/v1/prediction/{prediction_id}",
            headers=HEADERS
        ).json()
        
        if result["status"] == "success":
            return result
        elif result["status"] == "error":
            raise Exception(f"Prediction failed: {result}")
        
        time.sleep(1)  # Wait before polling again

3. Complete Example

Here's a complete example that puts it all together, including error handling and result processing. This shows how to create a prediction and wait for the result in a production environment.

try:
    # Create prediction
    prediction_id = create_prediction()
    print(f"Prediction created: {prediction_id}")
    
    # Get result
    result = get_prediction(prediction_id)
    print(f"Output URL: {result['output']}")
    print(f"Processing time: {result['metrics']['predict_time']}s")
except Exception as e:
    print(f"Error: {e}")

Additional Information

The API uses a two-step process: create prediction and poll for results
Response time: ~261 seconds
Rate limit: 60 requests/minute
Concurrent requests: 10 maximum
Use long-polling to check prediction status until completion

Overview

Mochi-1 is a state-of-the-art text-to-video generation model designed to create high-quality, dynamic videos from textual descriptions. By leveraging advanced machine learning techniques, it bridges the gap between creativity and technology, enabling users to transform their imagination into seamless, visually appealing video outputs.

Technical Specifications

Mochi-1 Video Generation: Converts text prompts into video sequences with smooth transitions and high-quality visuals.
Adaptive Scaling: Ensures outputs are consistent across various frame rates and resolutions.
Input Flexibility: Accepts detailed parameter configurations to provide users with extensive control over video characteristics.
Seed Control: Reproducibility through seed values enables consistent outputs for the same parameter settings.

Key Considerations

Frame Count Limitations: Mochi-1 supports a range of 30-170 frames. Exceeding these limits may result in errors or degraded performance.
Frame Rate (FPS): Set between 10-60 FPS for smooth playback. Higher FPS values require additional computational power.
Guidance Scale: Ranges from 1 to 10, controlling the adherence to the textual prompt. Extreme values may reduce output quality.
Prompt Strength: Adjusted between 0-1, impacting the influence of image-based prompts relative to text.
Seed Consistency: The seed value determines output reproducibility. Keep it consistent for identical results across runs.

Tips & Tricks

Optimal Frame Count (num_frames):
- Use 30-70 for short, concise clips.
- Set 100-150 for extended, detailed sequences.
- Avoid the maximum (170) unless necessary, as it may increase generation time significantly.
Image Prompt Strength (image_prompt_strength):
- Set 0.3-0.5 for balanced text and image influence.
- Use higher values (0.6-0.8) for image-dominant outputs.
- Avoid 1 unless text input is minimal, as it may overpower the text prompt.
Guidance Scale (guidance_scale):
- Use 3-5 for natural and balanced outputs.
- Set 6-8 for stronger alignment with text prompts.
- Avoid extreme values (1 or 10) as they may reduce coherence.
Frame Rate (fps):
- Use 24-30 for cinematic quality.
- Set 40-50 for dynamic or fast-paced visuals.
- Avoid 60 FPS unless necessary for specific use cases, as it increases computational load.
Seed (seed):
- Use a fixed value for reproducibility.

Capabilities

Text-to-Video: Mochi-1 converts descriptive text into high-quality video clips.
Customizable Parameters: Provides extensive control over frame count, prompt strength, FPS, and more.
Reproducibility: Seed control enables consistent outputs for the same configuration.
Dynamic Visuals: Smooth transitions and coherent sequences.

What can I use for?

Creative Projects: Mochi-1 generates videos for storytelling, marketing, and design.
Prototyping: Rapidly visualize concepts or ideas.
Education: Create visual aids and demonstrations.
Entertainment: Produce visually appealing clips for social media or personal use.

Things to be aware of

Creative Storytelling: Use vivid and imaginative prompts to craft compelling narratives.
Dynamic Compositions: Experiment with various FPS and frame counts to suit different styles.
Prompt Strength Balance: Adjust the image and text prompt strengths for hybrid inspirations.
Reproducibility: Use a fixed seed to iterate on a consistent baseline.

Limitations

Prompt Sensitivity: Ambiguous or overly complex prompts may result in inconsistent outputs.
Balance Challenge: Finding the ideal parameter configuration may require multiple iterations.
Output Consistency: While seeds ensure reproducibility, varying parameter combinations may lead to unexpected results.

Output Format: MP4

Related AI Models

Pyramid Flow

pyramid-flow

Text to Video

Wan 2.1-1.3B

wan-2-1-1-3b

Text to Video

Gen-2 by Runway

runway

Text to Video

Video Crafter

video-crafter

Text to Video