Soutine AI

Soutine AI is a freemium, multi-model image and video generation studio combining prompt libraries, shared history, and credit-based access to text-to-image, image-to-image, text-to-video, image-to-video, and upscaling tools.

Soutine AI is a web-based studio that brings several image and video generation models together under one roof, letting users move from a text prompt or reference image to a finished visual without juggling separate tools. The workspace supports text-to-image, image-to-image, text-to-video, image-to-video, and video upscaling, with models such as Qwen Image 2.1, Nano Banana, Nano Banana 2 Lite, Nano Banana Pro, GPT Image 2.5, and Kling 4 selectable per task. The product targets people who want visual output quickly and without prior editing skill: marketers building ad creative, social content creators, illustrators looking for reference material, and anyone experimenting with AI-generated imagery. Soutine AI keeps every generated asset in a shared history called My Creations, so a first frame produced from a prompt can later be sent into video generation or upscaling without restarting the project. Curated prompt libraries for Nano Banana and Qwen Image 2.1, complete with before and after examples, give newcomers a starting point rather than a blank text box. Access runs on a credit system. New accounts get free credits after signing up with Google, with no card required, and can generate on the free plan where outputs stay public and carry a watermark. Paid membership tiers raise monthly credit allowances, unlock private and watermark-free generation, add commercial-use rights, and increase concurrent generation slots and queue priority. Credit cost per generation varies by model, resolution, and duration, and failed system or provider tasks are refunded automatically. Within the broader field of AI creative tools, Soutine AI positions itself as a multi-model aggregator rather than a single proprietary model. Its value lies in consolidating several leading image and video generators, their prompt libraries, and generation history into one interface, rather than in offering a model built in-house.