# AI models available on PicVideo > Content revision: 2026-10-17 Each starting cost below uses the cheapest setting a model offers; your exact price for the settings you pick appears in the workspace before you generate. Models we highlight first: Seedance 2.5, MiniMax H3, Veo 3.1 Lite, Kling 3.0, Nano Banana Pro, GPT Image 2.5, Seedream 5 and Qwen Image 3. ## Video models ### Runway Gen-4 Turbo Turn a still photo or a short written idea into a 5- or 10-second clip in five frame shapes, with 1080p kept for the 5-second length. - Modes: Text to video and Image to video - Resolutions: 720p and 1080p - Durations: 5 and 10 seconds - Sound: Not supported - From 24 credits per video ### Gemini Omni Flash 1.1 Pin a first and last frame from two photos, or mix image, video and audio references, for 4- to 10-second clips from 360p up to 4K. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 360p, 720p, 1080p and 4K - Durations: 4, 6, 8 and 10 seconds - Sound: Not supported - From 90 credits per video ### Gemini Omni Bring character photos and short reference clips into one widescreen or vertical scene of 4 to 10 seconds at 720p, 1080p or 4K. - Modes: Text to video, Image to video and Reference to video - Resolutions: 720p, 1080p and 4K - Durations: 4, 6, 8 and 10 seconds - Sound: Not supported - From 90 credits per video ### Veo 3.1 Premium The quality tier of Google’s Veo 3.1 for a polished portrait or product shot that comes back with sound, in 4, 6 or 8 seconds up to 4K. - Modes: Text to video and Image to video - Resolutions: 720p, 1080p and 4K - Durations: 4, 6 and 8 seconds - Sound: Included in every video - From 450 credits per video ### Veo 3.1 Basic A quicker Veo 3.1 tier than Premium that, like Lite, also takes a first-and-last-frame pair or reference photos, with built-in sound in 4, 6 or 8 seconds. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 720p, 1080p and 4K - Durations: 4, 6 and 8 seconds - Sound: Included in every video - From 60 credits per video ### Veo 3.1 Lite The lightest Veo 3.1 tier still returns every clip with sound, which makes it handy for trying several takes of a family snapshot or travel photo. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 720p, 1080p and 4K - Durations: 4, 6 and 8 seconds - Sound: Included in every video - From 30 credits per video ### Seedance 2.5 Animate a portrait or product photo for up to 30 seconds, steering it with first and last frames, reference clips or audio cues; clips come back silent for your own soundtrack. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 480p, 720p and 1080p - Durations: 4 to 30 seconds - Sound: Not offered on PicVideo (clips are silent) - From 224 credits per video ### Seedance 2.0 ByteDance’s Seedance 2.0 combines photos, video clips and audio cues into 4- to 15-second scenes, from 480p up to 4K and ultra-wide 21:9. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 480p, 720p, 1080p and 4K - Durations: 4 to 15 seconds - Sound: Optional - From 152 credits per video ### Seedance 2.0 Fast A quicker Seedance 2.0 route at 480p or 720p for drafting pet clips or product moments from your photos before you pick a final take. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 480p and 720p - Durations: 4 to 15 seconds - Sound: Optional - From 94 credits per video ### Seedance 2.0 Mini The smallest Seedance 2.0 variant keeps keyframes, references and optional sound for short 480p or 720p social drafts from a single snapshot. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 480p and 720p - Durations: 4 to 15 seconds - Sound: Optional - From 31 credits per video ### HappyHorse 1.0 HappyHorse 1.0 turns a prompt, a photo or a reference image into a 3- to 15-second clip at 720p or 1080p. - Modes: Text to video, Image to video, Reference to video and Video to video - Resolutions: 720p and 1080p - Durations: 3 to 15 seconds - Sound: Not supported - From 168 credits per video ### HappyHorse 1.1 HappyHorse 1.1 adds more frame shapes, from 4:5 portrait to 21:9 cinema, for photo or reference-image clips of 3 to 15 seconds. - Modes: Text to video, Image to video and Reference to video - Resolutions: 720p and 1080p - Durations: 3 to 15 seconds - Sound: Not supported - From 135 credits per video ### Wan 3.0 Prime The Prime tier of Wan 3.0 suits longer photo stories of 2 to 30 seconds, guided by start and end frames or reference media, with optional sound. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 480p, 720p and 1080p - Durations: 2 to 30 seconds - Sound: Optional - From 49 credits per video ### Wan 3.0 Wan 3.0 at its standard tier builds 2- to 30-second clips from a photo, a start-and-end pair or mixed references, at 480p, 720p or 1080p. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 480p, 720p and 1080p - Durations: 2 to 30 seconds - Sound: Optional - From 32 credits per video ### Kling 3.0 Kling 3.0 can split one photo or prompt into several shots, with optional sound, any length from 3 to 15 seconds and output up to 4K. - Modes: Text to video and Image to video - Resolutions: 720p, 1080p and 4K - Durations: 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15 seconds - Sound: Optional - From 84 credits per video ### Kling 3.0 Turbo A faster Kling 3.0 route without an audio setting, giving 3- to 15-second clips at 720p or 1080p when a portrait just needs a quick motion test. - Modes: Text to video and Image to video - Resolutions: 720p and 1080p - Durations: 3 to 15 seconds - Sound: Not supported - From 108 credits per video ### Kling 2.6 Kling 2.6 makes 5- or 10-second 1080p clips from a photo or prompt and lets you decide whether generated sound is included. - Modes: Text to video and Image to video - Resolutions: 1080p - Durations: 5 and 10 seconds - Sound: Optional - From 110 credits per video ### Grok Imagine Grok Imagine from xAI turns a photo or prompt into a 6- to 30-second clip in portrait, landscape or square framing at up to 1080p. - Modes: Text to video and Image to video - Resolutions: 480p, 720p and 1080p - Durations: 6 to 30 seconds - Sound: Not supported - From 29 credits per video ### Grok Imagine Video 1.5 Grok Imagine Video 1.5 from xAI starts from your photo and returns a clip of 1 to 15 seconds at 480p, 720p or 1080p. - Modes: Image to video - Resolutions: 480p, 720p and 1080p - Durations: 1 to 15 seconds - Sound: Not supported - From 10 credits per video ### MiniMax H3 MiniMax H3 always delivers sound with the picture and reads first and last frames or mixed references, for 4- to 15-second clips at 768p or 2K. - Modes: Text to video, Image to video, Start and end frame and Reference to video - Resolutions: 768p and 2K - Durations: 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15 seconds - Sound: Included in every video - From 64 credits per video ### PixVerse V6 PixVerse V6 covers prompts, photos, frame pairs, references and clip extension, with a sound switch and lengths from 1 to 15 seconds. - Modes: Text to video, Image to video, Start and end frame, Reference to video and Video to video - Resolutions: 360p, 540p, 720p and 1080p - Durations: 1 to 15 seconds - Sound: Optional - From 16 credits per video ### Wan 2.7 Wan 2.7 animates a photo, bridges two keyframes or follows reference images and clips, at 720p or 1080p for 2 to 15 seconds. - Modes: Text to video, Image to video, Start and end frame, Reference to video and Video to video - Resolutions: 720p and 1080p - Durations: 2 to 15 seconds - Sound: Not supported - From 64 credits per video ### Kling 3.0 Omni Kling 3.0 Omni handles prompts, single photos and first-to-last-frame pairs, with custom multi-shot plans, optional audio and up to 4K. - Modes: Text to video, Image to video and Start and end frame - Resolutions: 720p, 1080p and 4K - Durations: 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15 seconds - Sound: Optional - From 84 credits per video ## Image models ### Nano Banana Pro Google’s Nano Banana Pro creates or edits pictures from up to eight reference photos, with 1K, 2K or 4K output for prints and detailed product shots. - Modes: Text to image and Image to image - Resolutions: 1K, 2K and 4K - From 16 credits per image ### Nano Banana 2 Nano Banana 2 from Google accepts up to 14 reference images and frames as extreme as 8:1, good for combining several family photos into one scene. - Modes: Text to image and Image to image - Resolutions: 1K, 2K and 4K - From 10 credits per image ### Nano Banana 2 Lite A lighter Google Nano Banana 2 option for quick photo edits and new pictures, taking up to 10 references across wide and tall frame shapes. - Modes: Text to image and Image to image - Resolutions: Model default - From 6 credits per image ### GPT Images 2.5 Sunburst OpenAI’s GPT Images 2.5 Sunburst is tuned for precision, a fit for careful edits from up to 16 references with output up to 4K. - Modes: Text to image and Image to image - Resolutions: 1K, 2K and 4K - From 6 credits per image ### GPT Images 2.5 Flare GPT Images 2.5 Flare is OpenAI’s quicker 2.5 option for restyling portraits or building new scenes, with up to 16 references and 1K to 4K output. - Modes: Text to image and Image to image - Resolutions: 1K, 2K and 4K - From 6 credits per image ### GPT Image 2 OpenAI’s GPT Image 2 spans frame shapes from 3:1 banners to 9:21 phone wallpapers and can return a transparent background for product cutouts. - Modes: Text to image and Image to image - Resolutions: 1K, 2K and 4K - From 6 credits per image ### Seedream 5.0 Pro ByteDance’s Seedream 5.0 Pro blends up to 10 reference photos into a new 1K or 2K image, useful for product sets or matching portraits. - Modes: Text to image and Image to image - Resolutions: 1K and 2K - From 14 credits per image ### Seedream 5.0 Flash Seedream 5.0 Flash from ByteDance creates or edits pictures at 1K, 1.5K or 2K, a practical everyday choice for photo touch-ups. - Modes: Text to image and Image to image - Resolutions: 1K, 1.5K and 2K - From 7 credits per image ### Seedream 5.0 Lite The Lite tier of ByteDance Seedream 5.0 reaches 3K output and takes up to 14 references for detailed edits of travel or wedding photos. - Modes: Text to image and Image to image - Resolutions: 1K, 2K and 3K - From 11 credits per image ### Qwen Image 2.1 Qwen Image 2.1 edits a photo you upload, with an optional mask to limit the change and a transparent or opaque background for product images. - Modes: Image to image - Resolutions: 1K and 2K - From 8 credits per image ### Grok Imagine Image 2.0 Grok Imagine Image 2.0 by xAI makes or edits pictures from up to five references in square, portrait or landscape frames. - Modes: Text to image and Image to image - Resolutions: Model default - From 8 credits per image ### Grok Imagine Grok Imagine from xAI returns two images for each text prompt and one when you edit an uploaded photo. - Modes: Text to image and Image to image - Resolutions: Model default - From 8 credits per image ### Flux Kontext Pro Flux Kontext Pro reads the scene in one uploaded photo before editing it, useful for changing the background or outfit in a portrait. - Modes: Text to image and Image to image - Resolutions: Model default - From 10 credits per image ### Flux Kontext Max Flux Kontext Max is the more detailed Kontext tier, meant for single-photo edits where small textures and fine features need care. - Modes: Text to image and Image to image - Resolutions: Model default - From 20 credits per image ### Z-Image Turbo Z-Image Turbo draws pictures from a written prompt only, and its developer says it can render bilingual text inside the image. - Modes: Text to image - Resolutions: Model default - From 2 credits per image ### Qwen Image 3.0 Qwen Image 3.0 writes new 1K or 2K pictures or edits up to three reference photos, with a negative prompt to keep unwanted details out. - Modes: Text to image and Image to image - Resolutions: 1K and 2K - From 10 credits per image ### Qwen Image 3 Pro The Pro tier of Qwen Image 3 handles the same prompt and three-reference edits at 1K or 2K for pictures you plan to publish. - Modes: Text to image and Image to image - Resolutions: 1K and 2K - From 13 credits per image ### Flux 2 Pro FLUX.2 Pro creates or edits images from up to eight references at 1K or 2K, a steady pick for consistent product or portrait series. - Modes: Text to image and Image to image - Resolutions: 1K and 2K - From 10 credits per image ### Flux 2 Flex FLUX.2 Flex is the adjustable sibling of FLUX.2 Pro, creating or editing images from up to eight references at 1K or 2K. - Modes: Text to image and Image to image - Resolutions: 1K and 2K - From 28 credits per image ### Imagen 4 Fast Imagen 4 Fast is Google’s quick prompt-only route for trying several picture ideas, such as a greeting card scene, before refining one. - Modes: Text to image - Resolutions: Model default - From 8 credits per image ### Imagen 4 Google’s Imagen 4 turns a written description into a new picture in five frame shapes, with a negative prompt and a seed for repeatable results. - Modes: Text to image - Resolutions: Model default - From 16 credits per image ### Imagen 4 Ultra Imagen 4 Ultra is the most detailed Imagen 4 tier from Google, for prompt-only pictures where fine texture and lighting matter. - Modes: Text to image - Resolutions: Model default - From 24 credits per image ## Official sources - [Pricing](https://picvideo.net/pricing)