Offline AI Image-to-Video Converter
Budget: $30 – $250 SGD
I need a Windows-based desktop application that turns a single still image into an AI-generated video, entirely offline. The workflow should rely on a local Stable Diffusion installation and incorporate the WAN2.6 model to drive the final animation—no external API calls, cloud processing, or telemetry of any kind.
Core requirements
• Runs natively on Windows 10/11 and installs with minimal user setup.
• Accepts at least JPEG files for input; adding PNG or BMP later should remain possible.
• Generates a short video (MP4 preferred) by feeding the image through Stable Diffusion and WAN2.6.
• Interface must feel intuitive for non-technical users while exposing advanced settings in an “expert” panel.
• Conversion speed is critical; please optimise GPU utilisation and let me choose device (CUDA / DirectML).
• Output parameters—resolution, frame rate, length, prompt text, CFG scale, seed—should all be editable before rendering.
Deliverables
1. Executable installer (or portable folder) with all weights and dependencies bundled for offline use.
2. Source code with clear build instructions so I can re-compile if models update.
3. Brief user guide covering installation, basic usage, and how to swap in updated checkpoints.
4. One sample project file proving the tool converts a JPEG into a smooth video using WAN2.6 locally.
Acceptance criteria
• I can disconnect the PC from the internet, open the app, load a JPEG, type a prompt, and export a video without errors.
• GPU memory footprint is documented and stays within practical limits for an 8 GB card.
• No calls are made to external servers (I will monitor network traffic during testing).
If you have deep experience with Stable Diffusion, animation extensions, and Windows desktop packaging, let’s discuss your proposed architecture and timeline.
Core requirements
• Runs natively on Windows 10/11 and installs with minimal user setup.
• Accepts at least JPEG files for input; adding PNG or BMP later should remain possible.
• Generates a short video (MP4 preferred) by feeding the image through Stable Diffusion and WAN2.6.
• Interface must feel intuitive for non-technical users while exposing advanced settings in an “expert” panel.
• Conversion speed is critical; please optimise GPU utilisation and let me choose device (CUDA / DirectML).
• Output parameters—resolution, frame rate, length, prompt text, CFG scale, seed—should all be editable before rendering.
Deliverables
1. Executable installer (or portable folder) with all weights and dependencies bundled for offline use.
2. Source code with clear build instructions so I can re-compile if models update.
3. Brief user guide covering installation, basic usage, and how to swap in updated checkpoints.
4. One sample project file proving the tool converts a JPEG into a smooth video using WAN2.6 locally.
Acceptance criteria
• I can disconnect the PC from the internet, open the app, load a JPEG, type a prompt, and export a video without errors.
• GPU memory footprint is documented and stays within practical limits for an 8 GB card.
• No calls are made to external servers (I will monitor network traffic during testing).
If you have deep experience with Stable Diffusion, animation extensions, and Windows desktop packaging, let’s discuss your proposed architecture and timeline.