PYRAMID-FLOW
Code of Pyramidal Flow Matching for Efficient Video Generative Modeling
Avg Run Time: 276.000s
Model Slug: pyramid-flow
Playground
Input
Enter a URL or choose a file from your computer.
Invalid URL.
image/jpeg, image/png, image/jpg, image/webp (Max 50MB)
Output
Example Result
Preview and download your result.
API & SDK
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.
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.
Readme
Overview
Pyramid Flow — Text to Video AI Model
Pyramid Flow is an open-source text-to-video AI model developed for researchers and organizations prioritizing efficient, responsible video generation. Rather than chasing maximum speed or resolution, Pyramid Flow balances quality with accessibility, making it ideal for research teams and ethical AI projects that need reliable video synthesis without enterprise-scale hardware requirements. The model supports both text-to-video and image-to-video workflows, enabling flexible input modalities for experimental video generation pipelines.
As an AI video generation model optimized for research applications, Pyramid Flow addresses a specific gap in the landscape: projects that require more capability than lightweight models but don't need the 80GB+ VRAM footprint of cinematic-grade systems. This positions it as a practical choice for academic institutions, independent researchers, and developers building AI video generation tools with ethical constraints.
Technical Specifications
What Sets Pyramid Flow Apart
Research-Focused Design Philosophy: Unlike commercial models optimized for rapid content creation, Pyramid Flow is explicitly architected for research and ethical AI projects. This means the model prioritizes reproducibility, transparency, and responsible development practices — critical for academic work and organizations with AI governance requirements.
Balanced Hardware Efficiency: Pyramid Flow requires 16GB minimum VRAM, positioning it as an efficient video AI model that delivers better quality than 8-12GB alternatives while remaining accessible to researchers without enterprise GPU clusters. This efficiency-to-quality ratio makes it practical for university labs and independent research teams.
Dual Input Modality Support: The model natively supports both text-to-video and image-to-video generation within a single framework. Researchers can feed either text prompts or reference images, enabling flexible experimentation without maintaining separate models for different input types.
Technical Specifications:
- Maximum resolution: Up to 720p
- Frame rate: 24 fps
- Supported inputs: Text prompts, reference images
- Minimum VRAM: 16GB
- Community support: Moderate with active research community
Key Considerations
The quality of the generated video is highly dependent on the clarity and relevance of the input prompts and images.
Longer durations may require more computational resources and could affect the coherence of the video.
Balancing the guidance scales is crucial to achieve the desired influence of text and image inputs on the final output
Tips & Tricks
How to Use Pyramid Flow on Eachlabs
Access Pyramid Flow through Eachlabs' Playground for interactive experimentation or integrate it via API for production workflows. Provide text prompts or reference images as input, specify your desired resolution (up to 720p) and frame rate (24 fps), and receive video outputs optimized for research and ethical applications. Eachlabs handles infrastructure, so you can focus on model experimentation without managing GPU resources.
Capabilities
Text-to-Video Generation with Pyramid Flow : Converts textual descriptions into dynamic video content.
Image-to-Video Generation: Transforms static images into animated sequences, guided by the provided image and optional text prompts.
Temporal Consistency: Maintains coherent motion and scene transitions across frames.
What Can I Use It For?
Use Cases for Pyramid Flow
Academic Research & Computer Vision Studies: Researchers studying video synthesis, generative models, or AI ethics can use Pyramid Flow to prototype and validate hypotheses without licensing restrictions. The open-source nature and research-grade architecture make it ideal for publishing reproducible results and contributing to the broader AI community.
Ethical AI Development & Governance: Organizations building responsible AI systems can integrate Pyramid Flow as a transparent, auditable video generation component. Teams developing AI governance frameworks or conducting bias/fairness studies benefit from the model's research-focused design and moderate community support for collaborative problem-solving.
Experimental Video Generation Pipelines: Developers prototyping novel video AI applications can leverage both text-to-video and image-to-video capabilities. For example, a developer might prompt: "A serene forest stream flowing over moss-covered rocks, morning mist rising, soft natural lighting" — then iterate on the output or feed reference images to explore how the model handles different input modalities.
Educational Content Creation: Educators and instructional designers can use Pyramid Flow to generate demonstration videos for computer science, AI, and media studies courses. The accessible hardware requirements and open-source availability make it practical for university labs teaching generative AI concepts.
Things to Be Aware Of
Experiment with different combinations of text prompts and images to discover unique video outputs.
Adjust the guidance scales to see how the influence of text and image inputs affects the generated content.
Vary the duration and frames per second to create videos with different pacing and styles.
Limitations
The Pyramid Flow may struggle with highly complex scenes or prompts that require intricate temporal dynamics.
There is a possibility of artifacts or inconsistencies in longer videos due to the challenges in maintaining coherence over extended durations.
The generated videos are limited by the diversity and quality of the data the Pyramid Flow was trained on.
Output Format:MP4
Pricing
Pricing Detail
This model runs at a cost of $0.001540 per second.
The average execution time is 276 seconds, but this may vary depending on your input data.
The average cost per run is $0.425040
Pricing Type: Execution Time
Cost Per Second means the total cost is calculated based on how long the model runs. Instead of paying a fixed fee per run, you are charged for every second the model is actively processing. This pricing method provides flexibility, especially for models with variable execution times, because you only pay for the actual time used.
Related AI Models
You can seamlessly integrate advanced AI capabilities into your applications without the hassle of managing complex infrastructure.
The most advanced video generation model by Google DeepMind. Creates realistic scenes, natural sounds, and physically consistent motion from a single text prompt. Perfect for storytelling, cinematic ads, and short films.
Create high resolution, long duration cinematic scenes faithful to your script by simply entering text prompts with minimax hailuo v2 3 pro text to video.
Cutting edge text to video generation delivering cinematic shots, lifelike motion dynamics, and seamless native audio all from a single prompt.
Generate cinematic videos with synchronized audio in seconds. The Fast mode of LTXV-2 delivers high-quality motion and sound at accelerated rendering speed
