Choosing the right model

There are a lot of AI models, and "which one should I use?" is genuinely hard -- it depends on what you are making, where you want to run it, and what your computer can handle. This guide walks the decision in that order. If you would rather talk it through, the AI Models page has a Help me choose button that opens the assistant, which asks about your project and hardware and recommends a specific model.

Start from how you plan to work

Before the goal, there is a question that narrows the field faster than anything else: how do you intend to make your shots? Answer it in your project's Techniques field (see Projects), because it decides which kind of model is even relevant.

• "I want to describe a shot and let the model animate it" points at text-to-video.
• "I want to approve a still, then bring it to life" points at image-to-video.
• "I want the shot to land on a specific pose" points at models that take a last frame as well as a first.
• "I want to act it out on camera and transfer the motion" points at video-to-video and pose control -- a different, heavier class of model.
• "I will record my own voice" points at voice cloning; "my characters are never shown talking on screen" means you can skip lipsync models entirely.

Two people making shows that look identical can need completely different models because they work differently. The assistant reads Techniques first for exactly this reason -- and if it is empty, it will ask before recommending anything.

Then start from your goal

Pick the kind of thing you are making; that narrows the field far more than any spec sheet:

• Still images -- character art, backgrounds, concept frames. Families like SDXL and Flux are common starting points; SDXL has a huge community of add-on styles (LoRAs), while Flux tends toward sharper, more literal results.
• Editing an image -- changing part of a picture, restyling, or combining references. These are separate "edit" models rather than the ones you use to make an image from scratch.
• Video -- turning an image or a prompt into a short clip. This is the most demanding kind of model. Video models differ a lot in speed and final quality: some are fast with a lively community of add-ons but softer results, others are slower but hold detail better. Because reports vary and change over time, treat any "X is better than Y" claim as a starting point, not a rule -- and lean on Comfy Cloud for video if your own card is modest.
• Upscaling -- making an existing image or frame larger and cleaner. These are small, specialised models.
• Voice and speech -- turning text into spoken dialogue, or cloning a voice. See the audio pages for how these fit the dialogue pipeline.

Model names get in the way here more than they should: a family name does not tell you what a model does, and a bigger version number does not always mean a newer model. If two candidates look like versions of the same thing, Model families and version names explains how to tell them apart.

Then decide where it runs

Every model can run one of two ways, and the choice changes what "which model" even means:

• Comfy Cloud -- the model runs on rented hardware. Nothing to install, and you always get the full-size model, but you pay for each render.
• Local ComfyUI -- the model runs on your own computer. Free to run once set up, and fully under your control, but you download multi-gigabyte files and your graphics card has to be big enough.

For the trade-off in detail, see Comfy Cloud or local?.

For local, your graphics card decides the version

Here is the part people miss: a single model is usually published in several versions that trade memory for quality and speed. A card with less memory can still run a big model -- you just use a smaller version of it. Which version fits your card is the whole game for local generation, and it is explained in Model versions and quantization.

You almost never have to choose the version by hand. When you run locally, the platform picks a build that fits your hardware; when you run on Comfy Cloud, it always uses the full-size file.

How the platform helps

• AI Models (a project view) is where you turn models on for a project. Turn on the ones you want; if your selection needs files you don't have locally, the page tells you so and sends you to the ComfyUI section of Settings, where the Files needed list shows exactly which files, how big they are, and which you already have -- so you only ever download what you actually chose.
• Help me choose opens the assistant, which reads your project and your saved hardware, recommends a model that fits, and can turn it on for you.
• The assistant is honest rather than salesy: it will tell you a model needs a big download, or will not fit your card, or costs money per render on the cloud.

Model quality, speed, and cost all move over time as new versions ship. Use these notes to get oriented, then check the model's own page (linked from its card) for the current picture.