{"pinia":{"locale":{"language":"en"},"ai":{"aiList":[{"value":"banana","label":"Nano Banana","scope":"image","active":true,"description":"The popular neural network from Google: Gemini Flash 3.0 Banana.

Send a text prompt or upload up to 14 images along with your request to use Remix mode."},{"value":"kling","label":"Kling","scope":"video","active":true,"description":"Access to all current models from the leading neural network Kling. Generate videos from text prompts, send up to 4 images for generation in Keyframes or Elements models. Full neural network settings and mode selection are available in the menu on the right."},{"value":"topaz_astra","label":"Topaz Astra","scope":"tool_video_upscaler","active":true,"description":null},{"value":"veo3","label":"Google Veo","scope":"video","active":true,"description":"Google's cutting-edge AI model. It can generate videos up to 720p and 8 seconds long with automatic sound and speech generation.

Send a text prompt or upload an image along with your request to start the generation."},{"value":"sora-images","label":"Sora (GPT) Image","scope":"image","active":true,"description":"

Create or edit your images in the top neural network Sora Images (also known as GPT Images) from the well-known company OpenAI.

Send a text prompt or upload up to 5 images along with your text request to use the Remix mode.

"},{"value":"seedream","label":"Seedream","scope":"image","active":true,"description":"

Seedream 5.0 (Pro) is ByteDance’s unified multimodal image model for text to image, image to image, style transfer, and editing. Improved visual reasoning follows complex prompts, keeps subjects consistent, and delivers more realistic results. Precise single reference style transfer and controllable edits help preserve untouched areas.

"},{"value":"seedance","label":"Seedance","scope":"video","active":true,"description":"

Seedance 2.0 is an advanced video generation model by TikTok, built for dynamic storytelling, smooth animation, and fast visual production. It supports text prompts and multimodal input: up to 6 images, 3 videos, and 2 audio files in a single generation flow.

Key capabilities:

image-to-video
video-to-video
audio-supported generation
multi-reference workflow
dynamic storytelling
fast processing

"},{"value":"seedream-editor","label":"Seedream 5 Pro Editor","scope":"tool_image_inpaint","active":true,"description":null},{"value":"seedance-editor","label":"Seedance 2.5 Editor","scope":"tool_video_inpaint","active":true,"description":null},{"value":"flux","label":"Flux","scope":"image","active":true,"description":"The Flux model set for image generation claims to be SOTA (state of the art), which is the status of the best in the industry. Choose a model and other settings in the right menu and send a text prompt to start."},{"value":"grok_video","label":"Grok Imagine","scope":"video","active":true,"description":"A top-tier video AI model by xAI for creating short clips (6–10 seconds) from a text prompt or an image. It excels at dynamic scenes, expressive camera moves, and cinematic atmosphere. Also can generate video with sound to make the result feel more lifelike. Send a text prompt or upload an image, then submit your request to start generating."},{"value":"runway-frames","label":"Runway Frames","scope":"image","active":true,"description":"Generate images in the popular Runway Frames neural network using text prompts with various unique styles or upload up to 3 images for reference. All settings are available in the menu on the right."},{"value":"beeble","label":"Beeble","scope":"video","active":true,"description":"

SwitchX changes the background, lighting, and environment in an existing video using a mask. It works best with calm, static shots without sudden scene changes, when you need to keep the subject, motion, and final result natural.ы

Key capabilities:

maskingds
background replacementa
lighting adjustment
motion preservation

"},{"value":"grok_image","label":"Grok Imagine","scope":"image","active":true,"description":"A conversational video generation model designed for fast, creative motion content directly within dialogue. Unlike traditional video models, it enables iterative scene refinement through natural language interaction, making it ideal for rapid concept testing, short social media videos, and experimental visual storytelling."},{"value":"runway","label":"Runway","scope":"video","active":true,"description":"Generate videos in the popular RunWay GEN-4 neural network. Turbo mode is available for use with a single image only. All settings are available in the menu on the right.

Send a text prompt or upload up to 3 images along with your request to bring your image to life or to work in multi-frame mode."},{"value":"sora","label":"Sora","scope":"video","active":true,"description":"Sora is an OpenAI model for generating videos, designed for creating videos from text or image. The model allows you to make it so that the object remains unchanged, even if it temporarily disappears from the field of view."},{"value":"luma_image","label":"Luma","scope":"image","active":true,"description":"A model designed for precise work with references. Luma Uni-1 understands style, composition, lighting, and details from uploaded images, so the results align with your vision rather than feeling randomly generated.\nSupports up to 4K resolution and up to 9 reference images.\nAdd your idea and references to create a cohesive, controlled, and visually strong result."},{"value":"happy_horse","label":"HappyHorse","scope":"video","active":true,"description":"HappyHorse is a visual AI model for bold, vibrant, and stylized images with a strong creative character. Unlike more realistic models, it focuses on artistic expression, unusual forms, and visual impact, making results feel freer and more experimental. A great choice for creative concepts, character designs, illustrations, social media content, and projects where style, mood, and originality matter most."},{"value":"higgsfield-soul","label":"Higgsfield Soul","scope":"image","active":true,"description":"TA highly aesthetic AI model that generates hyper-realistic photos with over 50 stylish presets in the spirit of 'fashion shooting without a camera'.

Allows training characters based on your own photos and generating images with their faces."},{"value":"ideogram","label":"Ideogram","scope":"image","active":true,"description":"Generate images from text or an image, change aspect ratios or the background of your image (keeping the main subject intact), use the remix function and the popular character reference mode."},{"value":"heygen","label":"HeyGen","scope":"video","active":true,"description":"One of the top neural networks for avatar generation. Upload an image and an audio file/voice message to get started"},{"value":"luma","label":"Luma","scope":"video","active":true,"description":"Luma: Dream Machine is one of the top video generation models today. The neural network can create high-quality realistic videos based on a text prompt or image in just a few minutes."},{"value":"wan_image","label":"Wan","scope":"image","active":true,"description":"A versatile line of generative models designed for consistent quality, smooth output, and reliable instruction interpretation. WAN models balance performance and stability, making them suitable for a wide range of creative tasks, from everyday content creation to more demanding visual projects."},{"value":"hailuo-minimax","label":"Hailuo Minimax","scope":"video","active":true,"description":"Hailuo MiniMax is MiniMax’s cinematic image model built for expressive, story-driven visuals. It stands out with dramatic composition, polished lighting, and a strong sense of mood, making generations feel closer to keyframes, posters, and visual concepts than standard AI images. Ideal for atmospheric portraits, film-style scenes, branded visuals, and creative experiments with a more emotional, high-impact look."},{"value":"higgsfield","label":"Higgsfield","scope":"video","active":true,"description":"Unique neural network that allows creating cinematic videos in excellent quality with precise control over camera movements, setting a new level of creative freedom for directors, clip makers, and content creators."},{"value":"wan_video","label":"Wan Video","scope":"video","active":true,"description":"A versatile line of generative models designed for consistent quality, smooth output, and reliable instruction interpretation. WAN models balance performance and stability, making them suitable for a wide range of creative tasks, from everyday content creation to more demanding visual projects."},{"value":"flux3_video","label":"FLUX 3 Video","scope":"video","active":true,"description":"FLUX 3 by Black Forest Labs — one general model that generates and edits video, with audio"},{"value":"banana_inpaint","label":"Nano Banana Pro Editor","scope":"tool_image_inpaint","active":true,"description":null},{"value":"clarity","label":"Clarity","scope":"tool_image_upscaler","active":true,"description":"Enhance image quality with smart upscaling that makes photos sharper, more detailed, and more expressive"},{"value":"elevenlabs","label":"Elevenlabs","scope":"audio","active":true,"description":"

ElevenLabs is a high-fidelity AI voice model for natural, expressive speech generation across languages. It stands out for strong voice consistency, lifelike prosody, and flexible control over style, emotion, and pacing, so narration and character voices feel less synthetic.

"},{"value":"magnific","label":"Magnific","scope":"tool_image_upscaler","active":true,"description":null},{"value":"topaz_ai","label":"Topaz AI","scope":"tool_video_upscaler","active":true,"description":null},{"value":"chatgpt","label":"ChatGPT","scope":"text","active":true,"description":null},{"value":"claude","label":"Claude","scope":"text","active":true,"description":"

A family of AI models by Anthropic for working with text, code, information analysis, and complex tasks. Excels at handling large amounts of context, following instructions accurately, and solving multi-step tasks.

"},{"value":"gemini","label":"Gemini","scope":"text","active":true,"description":null},{"value":"grok","label":"Grok","scope":"text","active":true,"description":null},{"value":"deepseek","label":"Deepseek","scope":"text","active":true,"description":null},{"value":"perplexity","label":"Perplexity","scope":"text","active":true,"description":null},{"value":"qwen","label":"Qwen","scope":"text","active":true,"description":"Qwen3-Max-Thinking is a flagship model built to handle complex tasks at scale, combining strong instruction following with improved reasoning and consistent answers."},{"value":"zai","label":"Zhipu AI","scope":"text","active":true,"description":null},{"value":"suno","label":"Suno","scope":"audio","active":true,"description":"Choose a generation mode (instrumental/by your words/automatic word generation using a language model), specify styles using presets or your own in the menu on the right, and send a text prompt to generate music in the most popular neural network for creating compositions."},{"value":"video2audio","label":"Video to audio","scope":"audio","active":true,"description":"Upload a video up to 20 seconds, and Syntx will add sounds to match every movement and every frame."}]},"ai-models":{"models":[]},"flags":{"flags":{}},"auth":{"token":null},"user":{"user":null,"plan":null,"planInfo":null,"userIdentities":[],"tokenHistoryItems":[],"tokenHistoryPagination":{"page":1,"pageSize":15,"total":0,"totalPages":0,"hasMore":false},"deletedAt":null,"showUserDeletedDialog":false},"settings":{"oauth":null,"ais":null,"ipInfo":null},"promo":{},"contest":{"contests":[],"contestsPage":1,"contestsPageSize":10,"contestsTotal":0,"activeContest":null,"contestGallery":[],"contestGalleryHasMore":true,"contestWinners":{},"contestStatus":{"uploaded":false}},"subscribe":{"subscribes":[],"tariffsFeatures":[],"isChangePlanModalVisible":false,"changePlanModalContext":null,"changePlanModalPlan":null}}}

Contests

A new contest is coming soon