How to make AI character-swap videos
Prepare two inputs, use a free or paid route and edit a mid-video switch. Includes the local graph and reference input.
Search
Search guides, questions and terms
Search the actual content on this site. Filter by type, or clear the search to browse everything.
29 entries
Build your first ComfyUI workflow
Follow the official purple-bottle recipe, inspect genuine publisher screenshots, and save an image with its reusable workflow.
How do I keep the same AI character across images?
Build a reference-editing baseline, change one scene at a time, and inspect identity before exporting. Local Qwen and paid Midjourney routes.
How to use a LoRA in ComfyUI
Match the adaptation to the base model, connect MODEL and CLIP correctly, then compare and export.
Why are details in my AI image prompt ignored?
Identify the missed requirement, distinguish wording from references and masks, then repair one region with free local inpainting or hosted editing.
How do I turn one picture into a short AI video?
Animate one authorised picture with the native Wan2.2 5B workflow, or compare Runway's hosted route. Includes exact files, motion prompt and export checks.
How to inpaint an image with a mask
Select the region you want to change, describe its replacement, and export a checked edit without discarding the original.
Does motion transfer also create lip sync?
Body movement and audio-driven mouth movement are separate tasks. Prepare a real speech input, use MuseTalk or Sync Labs, and check licences and export.
How do I turn a headshot into a cartoon and choose the clothes?
Make a recognisable cartoon baseline, then edit the outfit separately. Complete Firefly browser steps and a local Qwen alternative.
What is AI inference?
Inference is running a trained AI model on an input to produce an output. It is the computation that happens when you ask a model a question or generate an image.
Large language model (LLM)
An LLM is a machine-learning model trained on large amounts of language data to perform tasks such as generating, classifying or transforming text.
Token
A token is a unit of input or output processed by a language model. It can be a word fragment, punctuation or another encoded unit.
Context window
The context window is the amount of token context a model can use for a request, subject to the model and runtime configuration.
VRAM
VRAM is memory available to a graphics processor. Local AI uses accelerator memory for model data and working state.
Quantisation
Quantisation represents model numbers with lower precision to reduce storage or memory requirements, with possible changes to output quality.
Retrieval augmented generation (RAG)
RAG retrieves relevant material from an external source and supplies it to a model as context for generating an answer.
Hallucination
In generative AI, hallucination describes fabricated or unsupported content presented as though it were established information.
Diffusion
Diffusion models learn to reverse a noise process and can generate media through repeated refinement steps.
Checkpoint
A checkpoint is saved model state, usually including weights. In image workflows, the term often refers to the main model file selected by a loader.
LoRA
LoRA, or low-rank adaptation, adapts a model by training comparatively small update matrices while keeping the base weights frozen.
ComfyUI
ComfyUI is software for assembling and running AI media workflows through a graph of connected nodes.
Workflow
A workflow is a connected set of processing steps, their settings and inputs. In ComfyUI it is a node graph that can be saved and shared.
Prompt
A prompt is input used to guide a generative model, including instructions and relevant context. Some models also accept images or other media.
Image-to-video
Image-to-video generates video frames conditioned on one or more images, sometimes with text or other controls.
Training
Training adjusts model parameters using data and an optimisation process. Fine-tuning continues training from an existing trained model.
Can my PC run local AI?
Check the exact model, runtime, available memory and intended job. There is no single RAM or GPU number that makes every local AI model practical.
Why are ComfyUI nodes missing?
A workflow names node types that your current installation cannot load. Missing code, failed imports and missing model files need different fixes.
What do Q4, Q8, FP16 and GGUF mean?
Q4, Q8 and FP16 describe numerical representations; GGUF describes a model-file format. Neither label alone tells you quality, compatibility or total memory.
Why does my local AI model run slowly?
Separate model loading, input processing and output generation, then check placement and workload size before changing settings.
No matching entry yet. Try a shorter term or browse another content type.