Pick a video, image, audio, or text (LLM) model, see the price in USD, and call it through the API or MCP.
One Fuzion credit balance covers the web app, the API, and MCP. No separate top-up · 1 credit = ฿1
64 models
The latest Seedance video model, served directly by BytePlus. Generates 480p–1080p video from 4 to 30 seconds with audio, takes image and video references, and supports reference, keyframe, edit, and extend modes.
Seedance 2.0 from BytePlus. 480p up to 4K video from 5 to 15 seconds with audio, image references, and reference, keyframe, and extend modes.
The faster Seedance 2.0 from BytePlus. 480p–720p video from 5 to 15 seconds with audio, for quicker, lower-cost runs.
The smallest Seedance 2.0 from BytePlus. 480p–720p video from 5 to 15 seconds with audio at the lowest 2.0 price.
Kling 4.0 on Fuzion. 720p–1080p video from 3 to 30 seconds from text or reference images.
Kuaishou's Kling 4.0 Pro. 720p–1080p video from 3 to 30 seconds with audio, using image, video, and audio references, plus a keyframe mode with first and last frames.
The lower-cost Kling 4.0. 720p video from 3 to 20 seconds from image and video references, or from a first frame.
Kling 2.6 on Srv.2. 720p–1080p clips of 5 to 10 seconds with audio from an image.
Kling 3.0 on Srv.2. 720p up to 4K video from 3 to 15 seconds from text or an image.
Alibaba's Wan 3.0. 480p–1080p video from 2 to 30 seconds with audio, image, video, and audio references, and prompts up to 20,000 characters.
MiniMax H3. 768p up to 2K video from 5 to 15 seconds with audio and image references.
xAI's Grok Video 1.5. 480p–1080p video from 6 to 15 seconds with audio and multiple reference images.
Grok Video 1.5 frame mode: start from one image or from text only. 480p–1080p video up to 15 seconds with audio.
The lower-cost Grok Video 1.5. 480p–1080p video from 6 to 15 seconds with audio and one optional reference image.
Google Omni 1.2 Flash. 360p up to 4K video from 3 to 15 seconds with audio, image and video references, plus keyframe and extend modes.
Omni Flash 1.1 on Fuzion. 720p–1080p video from 4 to 10 seconds with image references.
Flux 3 served directly by Black Forest Labs. 720p–1080p video from 5 to 20 seconds with audio, image and video references, and a draft mode.
Vidu Q4. 540p–4K video from 3 to 16 seconds with audio, guided by up to 15 reference images and 3 voice references for consistent characters.
Google Omni Flash 1.1 on Srv.4. 720p–1080p video from 8 to 10 seconds with image references.
Veo 3.1 Fast on Fuzion. 720p up to 4K, 8-second video with up to 3 reference images.
Meta's Muse 1.0. 720p, 10-second video with up to 4 reference images.
Omni Flash 1.1. 720p–1080p video from 4 to 10 seconds with image references.
Seedream 5.5 is an AI image generation model from BytePlus on the Fuzion API. Supports 1K / 1.5K / 2K. Accepts image references.
BytePlus Seedream 5.0 Pro. Creates and edits 1K–2K images with up to 10 reference images, plus a layer separation mode.
The fast Seedream 5.0. Creates and edits 1K–2K images with up to 10 reference images.
GPT Image 2.6 Sunburst. Creates and edits 1K–4K images with up to 16 references, with GIF and SVG export.
GPT Image 2.6 Flare. Creates and edits 1K–4K images with up to 16 references, and can export an animated GIF or a vector SVG.
Midjourney v8.2. Artistic 1K–2K images with reference images.
GPT Image 2.5 Sunburst. Creates and edits 1K–4K images with up to 16 reference images.
Google's Nano Banana 2.1. Creates and edits 1K–4K images with up to 14 reference images.
Higgsfield's Soul. Fashion and portrait images at 720p–1080p with ready-made styles.
The ultra-fast Nano Banana 2. Low-cost 1K images with up to 14 reference images.
Google's Nano Banana 2. Creates and edits 1K–4K images with up to 14 reference images.
Google's Nano Banana Pro (Gemini 3 Pro Image). Creates and edits 1K–4K images with up to 14 reference images.
Recraft V4.1. 1K–2K images suited to graphics and illustration, with reference images.
Seedream 5.0 Lite. 2K–3K images with up to 14 reference images.
BytePlus Seed Audio. Audio from text up to 120 seconds, with audio and image references.
ElevenLabs v4 text-to-speech with several speakers in one file and voices from the library. Priced by character.
The fast ElevenLabs v4 text-to-speech. Priced by character.
Google's Gemini 3.8 Flash TTS. 24 kHz WAV speech with up to 2 speakers and line-by-line delivery direction.
Change the voice in a recording while keeping its timing. Files up to 5 minutes, priced by length.
Sound effects from an English description, 0.5 to 30 seconds long.
Transfer motion from a reference video to the character in an image. 720p–1080p, priced by clip length.
Kling Motion Control 2.6. Transfers motion from a video to the character in an image at 720p–1080p.
Kling Motion Control 3.0. Transfers motion from a video to the character in an image at 720p–1080p.
BytePlus OmniHuman 2.0. Makes a person in an image speak along with an audio file. 720p, as long as the audio, up to 5 minutes.
Kling Avatar makes a character image speak along with an audio file up to 2 minutes long.
BytePlus Fuklone swaps the character in a 4–30 second clip for one from your images while keeping the original motion and camera. 480p–1080p.
The lower-cost BytePlus Fuklone. Swaps the character in a clip at 720p–1080p, priced by clip length.
The fast BytePlus Fuklone. Swaps the character in a clip at 720p, priced by clip length.
GENJUTSU swaps the character in a video clip at 480p–1080p, priced per second of the source clip.
The lower-cost GENJUTSU at 480p–1080p, priced per second of the source clip.
DeepSeek V4.1 Flash is DeepSeek's fast text model with a 1M-token context and image input. Suited to agents, coding, and long-document analysis, with function calling and JSON output.
DeepSeek V4.1 Flash Economy is the lower-cost tier of V4.1 Flash, with a 1M-token context and image input. A good fit for high-volume work where speed matters less.
DeepSeek V4 Pro is DeepSeek's larger text model for work that needs deeper reasoning. 1M-token context, text input, with function calling and JSON output.
Dola Seed 2.1 Turbo is the latest BytePlus text model for general agent work. 256K-token context, text and image input, with reasoning, tool use, and structured output.
Seed 2.0 Lite is a BytePlus text model that balances price and quality. 256K-token context with text and image input; a higher rate applies to prompts over 128K tokens.
Seed 2.0 Mini is the smallest and cheapest BytePlus text model. 256K-token context with text and image input, suited to simple high-volume tasks; a higher rate applies to prompts over 128K tokens.
Fuzion Upscale is an AI upscaling model on the Fuzion API. Supports 1080p–4K. Accepts video references.
Topaz Video Upscale is an AI upscaling model from Topaz on the Fuzion API. Supports 720p–4K. Accepts video references.
BytePlus Enhance is an AI upscaling model on the Fuzion API. Supports 1080p–4K. Accepts video references.
Topaz Image Upscale is an AI upscaling model from Topaz on the Fuzion API. Accepts image references.
Black Forest Labs' FLUX Upscale enlarges video to 1080p, 2K, or 4K in two modes: Precise stays close to the source and Creative adds new detail. MP4 up to 20 seconds and 50MB.
