LTX-2.5 (Pro): FLF2V
Generate a video from two reference images using LTX-2.5, blending the content of both frames into a single animated sequence. Ideal for creating transition shots, morphing between two subjects or scenes, and building multi-shot narratives from still references.
About
This workflow turns two still images into a single, fluid video transition using the LTX-2.5 image-to-video API. Two LoadImage nodes supply your start and end reference frames, which are passed into LtxApi25ImageToVideo. The node blends structural cues from both inputs to synthesize an in-between sequence that morphs from the first image toward the second, producing a natural-looking transition, reveal, or morph. Finally, SaveVideo encodes the generated frames to a video file you can review or drop into an edit.
What you provide
- Reading the workflow
Runs in ComfyUI. This workflow can't run on this page. Open it in Comfy Cloud with everything loaded, or download it for your own ComfyUI.
Try in CloudQuestions & answers
It’s strongly recommended. The node or API may auto-resize, but mismatched sizes/aspects often cause stretching or warped motion. Align aspect ratio and approximate subject scale for the cleanest transition.
Use the duration and frame rate options in LtxApi25ImageToVideo to set total length and temporal resolution. If the node exposes blend or transition controls, adjust them to favor a gentler or more aggressive move between frames. Longer durations and consistent subject framing usually look smoother.
SaveVideo writes the generated sequence to a file path you set in the node (ComfyUI’s output directory by default). Choose your target frame rate there; the exact container/codec options depend on your environment and SaveVideo configuration.
Yes. LtxApi25ImageToVideo uses an API to run LTX-2.5 remotely, so your images are uploaded for processing. Ensure you’ve configured your API credentials and reviewed the provider’s privacy and data-retention policies before use.