Segmind

segmind-anime-video

Generate anime, manga, or cartoon-styled video with Segmind — a cel-shaded clip from a prompt, an anime-style ad, a stylised animation from a reference image or character. Use when the user asks for anime, manga, cartoon, cel-shaded, or hand-drawn animation as the *look* of a video. Do not use for photoreal video (that is segmind-generate-video), for still anime images, or for restyling existing footage.

segmind-faceswap

Swap a face into a video or an image with Segmind, using a reference photo of the face to put in. Use when the user wants someone's face put onto footage or a picture, a character recast, a cameo inserted, or a face replaced — including as a follow-on step after upscaling or generating a clip. Do not use for changing an expression, for reposing or restyling a person, for generating a talking avatar from a photo, or for anything involving a real person who has not consented.

segmind-fashion

Produce fashion and apparel imagery with Segmind — virtual try-on, flat-lay and packshot photography, mannequin-to-model, model or fabric swaps, lookbook and e-commerce product shots, and fashion showcase video. Use when the subject is clothing, accessories, or a person wearing them for a catalogue, storefront, lookbook, or ad. Do not use for general product photography of non-apparel goods, for portrait retouching unrelated to garments, or for face swapping (that is segmind-faceswap).

segmind-generate-image

Generate or edit a still image with Segmind. Use when the user wants a picture made from a prompt — a product shot, poster, logo, illustration, concept art, ad creative, thumbnail — or wants an existing image changed: restyled, relit, background swapped or removed, an object erased or added, text corrected, a face swapped, or upscaled. Also use for a series of related images (variations, a set of ad sizes, a storyboard's frames). Do not use for video or animation, for audio, or when the user only wants advice on which model to use.

segmind-generate-video

Generate or edit video with Segmind. Use when the user wants a clip made from a prompt or from a still image — a product or brand ad, a cinematic shot, an animated photo, a talking avatar, a reference-consistent character clip — or wants an existing video changed: extended, restyled, background removed, objects erased, lip-synced, or subtitled. Do not use for still images, for audio-only output, or when the user only wants a model recommendation.

segmind-pick-model

Choose a Segmind model, or answer a question about the catalog, before running anything. Use when the user asks which model to use, what a generation will cost, what is cheapest or fastest for a task, how many models exist, what Segmind supports for some modality, or names a capability without naming a model ("something that does virtual try-on"). Also use as the first step of any generation request where no model is pinned. Do not use when the user already named a slug and just wants it run.

segmind-video-edit

Edit, convert, or caption an existing video with Segmind — trim, cut, crop, resize, reformat for a platform, change speed, extract frames or audio, make a GIF, concatenate clips, re-encode, or burn in subtitles. Use when the user wants mechanical changes to footage they already have rather than new generated content. Do not use for generating video, for changing what is *in* the frame (that is a video-to-video model), or for upscaling resolution (that is segmind-video-upscale).

segmind-video-upscale

Upscale or enhance the resolution of an existing video with Segmind. Use when the user wants a video made sharper, bigger, higher-resolution, 2x/4K, cleaned up, or "enhanced" — including a clip they just generated that came out too soft. Do not use for images (that is segmind-generate-image), for generating new video, or for reformatting, trimming, or re-encoding a video without changing its resolution.