# AI models available on Flow AI > Content revision: 2026-11-09 Starting credit costs use the settings noted below and exclude promotional or subscription discounts. Your model, references, quality and length can change the final cost. Music and audio prices are per generation request. Choose from 17 video and 13 image models, plus Music V6 for music and Seed Audio 1.0 for audio. ## Video models ### Veo 3.1 Free Turn a written scene or photo into a short video, or connect a first and last frame to direct a transition. Image-reference mode accepts up to three pictures to guide the scene. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Credits: None required - Access: No sign-up required - Downloads: No watermarks - Speed: Slower generation - [Start creating: Veo 3.1 Free](https://flowai.im/create/video?model=veo-3.1-free) ### Veo 3.1 Lite Draft a product reveal or a scene from your own photos with the lowest-credit Veo tier. Use a first-and-last-frame pair to direct a transition or up to three reference pictures to guide the scene. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Sound: Native audio generation - Enhancement: Optional 1080p upgrade for 16:9 videos - From 25 credits per video - [Start creating: Veo 3.1 Lite](https://flowai.im/create/video?model=veo3_lite) ### Veo 3.1 Basic Use the everyday Veo tier for a portrait, a product shot or a written scene. Two keyframes can steer the beginning and ending, while reference mode accepts up to three images. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Sound: Native audio generation - Enhancement: Optional 1080p upgrade for 16:9 videos - From 50 credits per video - [Start creating: Veo 3.1 Basic](https://flowai.im/create/video?model=veo3_basic) ### Veo 3.1 Premium Choose the premium Veo tier for a scene you want to develop from text or photographs. Frame pairs and image references let you describe the action while supplying the visual material to follow. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Sound: Native audio generation - Enhancement: Optional 1080p upgrade for 16:9 videos - From 200 credits per video - [Start creating: Veo 3.1 Premium](https://flowai.im/create/video?model=veo3_premium) ### Seedance 2.5 Build a longer product sequence or character scene with frame pairs and mixed references. Reference mode accepts up to 30 images, 10 video clips and 10 audio files, subject to duration limits. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Resolutions: 480p and 720p - Durations: 4–30 seconds - Sound: Optional generated audio - From 120 credits per video (4-second output at 480p, with one 2-second video reference; text-only starts at 132 credits) - [Start creating: Seedance 2.5](https://flowai.im/create/video?model=seedance-2.5) ### Seedance 2.0 Develop a scene from a brief, animate a photograph or guide a new shot with reference media. The standard model offers 1080p as well as smaller drafts, with up to nine images, three videos and three audio references. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Resolutions: 480p, 720p and 1080p - Durations: 4–15 seconds - Sound: Optional generated audio - From 78 credits per video (4-second output at 480p, with one 2-second video reference; text-only starts at 80 credits) - [Start creating: Seedance 2.0](https://flowai.im/create/video?model=seedance-2.0) ### Seedance 2.0 Fast Try short versions of a social clip or product shot with the Fast variant. It keeps frame-pair and mixed-reference workflows at 480p or 720p, with up to nine images, three videos and three audio references. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Resolutions: 480p and 720p - Durations: 4–15 seconds - Sound: Optional generated audio - From 60 credits per video (4-second 480p text-only output, sound off) - [Start creating: Seedance 2.0 Fast](https://flowai.im/create/video?model=seedance-2.0-fast) ### Seedance 2.0 Mini Use the Mini variant to test a motion idea at 480p or 720p before expanding it. Text, image, frame-pair and mixed-reference modes are available, with up to nine images, three videos and three audio references. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Resolutions: 480p and 720p - Durations: 4–15 seconds - Sound: Optional generated audio - From 60 credits per video (4-second 480p text-only output, sound off) - [Start creating: Seedance 2.0 Mini](https://flowai.im/create/video?model=seedance-2.0-mini) ### MiniMax H3 Create a 2K scene from a written brief, a first-and-last-frame pair or mixed media. Reference mode accepts up to nine images, three videos and three audio files for a subject or scene you want to guide. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Resolutions: 2K - Durations: 4–15 seconds - From 120 credits per video (4-second 2K text-only output) - [Start creating: MiniMax H3](https://flowai.im/create/video?model=minimaxh3) ### Google Omni Lite Start a video from text, a photograph or a pair of frames using account credits without a subscription or credit-pack purchase. Use up to three image references when you want to guide the scene from several pictures. - Modes: Text to video, Image to video, Start and end frame, Reference to video - Enhancement: Optional 1080p upgrade for 16:9 videos - From 25 credits per video - [Start creating: Google Omni Lite](https://flowai.im/create/video?model=googleomnilite) ### Google Omni Turn a written scene or one source image into a short landscape or vertical clip. Reference mode combines one to three images with an optional video reference, with output choices up to 4K. - Modes: Text to video, Image to video, Reference to video - Resolutions: 720p, 1080p and 4K - Durations: 4, 6, 8 or 10 seconds - From 60 credits per video (4-second 720p text-only output) - [Start creating: Google Omni](https://flowai.im/create/video?model=googleomni) ### HappyHorse 1.0 Animate one photo or create a new scene using up to nine image references. Choose a short duration and a portrait, landscape or square composition for a product shot or visual story. - Modes: Text to video, Image to video, Reference to video - Resolutions: 720p and 1080p - Durations: 3–15 seconds - From 135 credits per video (3-second 720p text-only output) - [Start creating: HappyHorse 1.0](https://flowai.im/create/video?model=happyhorse) ### Kling 3.0 Make a short scene from text or use up to two images to guide its frames. Standard and Pro are quality settings within this model; choose whether to generate sound with the clip. - Modes: Text to video, Image to video, Start and end frame - Durations: 5, 10 or 15 seconds - Sound: Optional; enabled by default - From 150 credits per video (5-second Standard output, sound off; Pro starts at 200 credits) - [Start creating: Kling 3.0](https://flowai.im/create/video?model=kling30) ### Wan 2.6 Create a short video from a prompt or animate one source photo. Choose 720p or 1080p and a 5-, 10- or 15-second length for a product moment, portrait or storyboard shot. - Modes: Text to video, Image to video - Resolutions: 720p and 1080p - Durations: 5, 10 or 15 seconds - From 100 credits per video (5-second 720p output) - [Start creating: Wan 2.6](https://flowai.im/create/video?model=wan26) ### Grok Turn a written scene or a single image into a video using the Normal, Fun or Spicy creative mode. Text-to-video offers square, portrait and landscape framing. - Modes: Text to video, Image to video - From 50 credits per video - [Start creating: Grok](https://flowai.im/create/video?model=grok) ### Kling 2.6 Animate one photograph or build a short scene from text with Kling 2.6. Choose a 5- or 10-second clip and turn generated sound on or off for the take. - Modes: Text to video, Image to video - Durations: 5 or 10 seconds - Sound: Optional; disabled by default - From 100 credits per video (5-second output, sound off) - [Start creating: Kling 2.6](https://flowai.im/create/video?model=kling) ### Wan 2.5 Make a 5- or 10-second clip from text or one source photo. Choose 720p or 1080p for a small product movement, portrait animation or scene draft. - Modes: Text to video, Image to video - Resolutions: 720p and 1080p - Durations: 5 or 10 seconds - From 100 credits per video (5-second 720p output) - [Start creating: Wan 2.5](https://flowai.im/create/video?model=wan25) ## Image models ### Nano Banana 2 Free Try a product composition, illustration or photo edit with a prompt and up to 14 reference images. Choose from five portrait, landscape and square aspect ratios to frame the result. - Modes: Text to image, Image to image - Resolutions: 1K - Credits: None required - Access: No sign-up required - Downloads: No watermarks - Speed: Slower generation - [Start creating: Nano Banana 2 Free](https://flowai.im/create/image?model=nano-banana-2-free) ### Nano Banana 2 Combine up to 14 reference pictures into a new composition or generate an image from text. Choose 1K, 2K or 4K and framing from narrow portraits to wide banners; 2K and 4K require a subscription or credit-pack purchase. - Modes: Text to image, Image to image - Resolutions: 1K, 2K, 4K - From 10 credits per image (1K) - [Start creating: Nano Banana 2](https://flowai.im/create/image?model=nano-banana-2) ### Nano Banana Pro Use up to eight photos to guide a new image or a detailed edit. The Pro model offers 1K, 2K and 4K output for product imagery, invitations or larger-format layouts. - Modes: Text to image, Image to image - Resolutions: 1K, 2K, 4K - From 12 credits per image (1K) - [Start creating: Nano Banana Pro](https://flowai.im/create/image?model=nano-banana-pro) ### Nano Banana Restyle a picture or assemble a scene from up to ten reference images. Text-only creation and a choice of portrait, landscape and square framing are also available. - Modes: Text to image, Image to image - From 5 credits per image - [Start creating: Nano Banana](https://flowai.im/create/image?model=nano-banana) ### GPT Image 2 Write a visual brief or supply up to 16 reference pictures for an edit. Choose one of five aspect ratios for a portrait, product composition or social image. - Modes: Text to image, Image to image - From 5 credits per image - [Start creating: GPT Image 2](https://flowai.im/create/image?model=gpt-image-2) ### Seedream 5 Create from text or bring up to 14 reference images into an edit. Choose Basic or High quality to develop a product layout or a coordinated set of visual ideas. - Modes: Text to image, Image to image - From 12 credits per image (Basic quality) - [Start creating: Seedream 5](https://flowai.im/create/image?model=seedream-5) ### Seedream 4.5 Generate a picture from a description or edit from up to 14 reference images. Choose Basic or High quality for visual drafts and follow-up revisions. - Modes: Text to image, Image to image - From 12 credits per image (Basic quality) - [Start creating: Seedream 4.5](https://flowai.im/create/image?model=seedream-4.5) ### Flux 2 Flex Create or edit from up to eight reference images at 1K or 2K. Use the same framing across a series of product concepts or choose automatic framing when editing. - Modes: Text to image, Image to image - Resolutions: 1K, 2K - From 25 credits per image (1K) - [Start creating: Flux 2 Flex](https://flowai.im/create/image?model=flux-2-flex) ### Flux 2 Pro Build a 1K or 2K image from text or up to eight references. Portrait, square and landscape choices make it suitable for product compositions and matching creative variations. - Modes: Text to image, Image to image - Resolutions: 1K, 2K - From 10 credits per image (1K) - [Start creating: Flux 2 Pro](https://flowai.im/create/image?model=flux-2-pro) ### Wan 2.7 Generate a new image or edit from up to nine references with Wan 2.7. This is an image model, offering 1K and 2K output; 2K requires a subscription or credit-pack purchase. - Modes: Text to image, Image to image - Resolutions: 1K, 2K - From 10 credits per image (1K) - [Start creating: Wan 2.7](https://flowai.im/create/image?model=wan-2.7) ### Wan 2.7 Pro Use the Pro image variant for new scenes or edits from up to nine references. Text-to-image offers 1K, 2K and 4K; reference edits offer 1K and 2K. - Modes: Text to image, Image to image - Resolutions: Text to image: 1K, 2K and 4K; Image to image: 1K and 2K - From 20 credits per image (1K) - [Start creating: Wan 2.7 Pro](https://flowai.im/create/image?model=wan-2.7-pro) ### Grok Imagine Create a picture from a prompt or restyle one source image. Choose square, 2:3 portrait or 3:2 landscape framing when creating from text, or use a photo as the starting point for an edit. - Modes: Text to image, Image to image - From 10 credits per image - [Start creating: Grok Imagine](https://flowai.im/create/image?model=grok-imagine) ### Z-Image Create a new picture from text without uploading a reference. Five aspect ratios suit a portrait, square card or landscape illustration. - Modes: Text to image - From 5 credits per image - [Start creating: Z-Image](https://flowai.im/create/image?model=z-image) ## Music models ### Music V6 Describe a genre, mood and arrangement to make a song or an instrumental, or supply an audio reference for a cover. Add your own lyrics and adjust vocal preference, style strength and creativity to shape the result. - Modes: Text to music and Audio to music - Durations: 60, 180 or 300 seconds in the compact composer; 10–360 seconds in the detailed composer - References: One audio file, 6–300 seconds, up to 50 MiB - Sound: Vocals or instrumental output - From 20 credits per generation request; a request can return multiple tracks - [Start creating: Music V6](https://flowai.im/create/music) ## Audio models ### Seed Audio 1.0 Write a voiceover, a conversation or a complete sound scene with dialogue and ambience. Up to three audio references can guide the result, with pitch, speed and volume controls for further direction. - Modes: Text to audio, with optional audio references - Durations: Follows the script and reference audio - References: Up to 3 files, each at most 30 seconds and 10 MiB - Sound: Speech, dialogue and sound-scene output - From 3 credits per request; audio-reference duration is charged at 1 credit per second, with a 3-credit minimum - [Start creating: Seed Audio 1.0](https://flowai.im/create/audio) ## Using reference media Choose references that fit your selected model and mode. The limits below describe the materials you can upload to guide a generation. - **Seedance 2.0 / Fast / Mini:** In reference mode, include at least one image or video; audio alone is insufficient. Up to 9 images, 3 videos and 3 audio files are accepted. Each video/audio reference must be 2–15 seconds, with at most 15 seconds total for videos and separately 15 seconds total for audio. - **Seedance 2.5:** Reference mode also needs an image or video, even when audio is supplied. Up to 30 images, 10 videos and 10 audio files are accepted. Each video/audio reference must be 2–30 seconds, with a 30-second total per media type. - **MiniMax H3:** Frame mode accepts one or two images. Reference mode needs at least one media input, within its limits of 9 images, 3 videos and 3 audio files. - **Google Omni:** Image-to-video takes one image. Reference mode takes 1–3 images and optionally one video; a video alone does not satisfy the image requirement. - **HappyHorse 1.0:** Image mode takes one picture; reference mode takes 1–9 images. Use photos to guide this model's reference mode. - **Veo choices and Google Omni Lite:** Use frame mode for up to two images or image-reference mode for up to three. Choose image references for these modes. Reference audio can guide the scene or performance. To request generated sound, choose a model and sound options suited to that result. ## Choosing a model Match the model's modes and reference support to the material you have, then compare output options and starting costs. Follow its Start creating link, choose the mode and add your prompt or media. For help matching a task to a tool, see [Choosing a workflow](https://flowai.im/capabilities.md#choosing-a-workflow). Check [login and plan requirements](https://flowai.im/pricing.md#access-conditions) before starting credit-based creation. ## Official sources - [Video workspace](https://flowai.im/create/video) - [Image workspace](https://flowai.im/create/image) - [Music workspace](https://flowai.im/create/music) - [Audio workspace](https://flowai.im/create/audio) - [Pricing](https://flowai.im/pricing.md)