New AI models arrive every day. Summon's Model Manager lets you add most new fal.ai models yourself, often the day they appear, without updating or reinstalling the panel. Search fal.ai's catalog, click Add, and the model appears in the matching tab, ready to generate.
✕ removes a model you don't use, and a removed built-in model can be added back through search at any time. In Options, you can sort the model menus by name, price or recently added.
Refresh from fal.ai re-reads the capabilities of your added models, so new options show up in the panel.
Some models need a control or an input the panel doesn't have yet, or return a result it can't show yet. The panel tells you when a model can't be added, and support for it arrives with a panel update.
Summon ships with 44+ curated models, including Nano Banana Pro, GPT Image 2.5, Seedream 5.0 and FLUX.2 [pro] for images; Seedance 2.5, Kling v3 Pro, Veo 3.1 Fast, MiniMax H3 and Ray 3.2 for video; Kling O3 Edit Video and LTX-2.3 Reframe for editing video; and sync-3, HeyGen v3 Precision and OmniHuman 1.5 for lipsync. The Model Manager adds to that list.
Added models run on your own fal.ai key, like the built-in ones. fal.ai bills them at its listed prices, and Summon adds no markup. For a model you haven't run yet, the estimate may read "unknown"; the real cost shows after the run.
The full tutorial ends by adding a Qwen image-editing model and using it straight away.