Sber's GigaChat has introduced new image generation and editing features, but there is confusion regarding which AI models are powering these capabilities. While official documentation and courses suggest GigaChat uses the older Kandinsky 3.1 model for image generation with limitations on editing, recent announcements for Kandinsky 6.0 Image describe advanced editing features like in-painting directly within the chat interface. Independent media reports suggest these Kandinsky 6.0 features are indeed integrated into GigaChat, offering unlimited image generations without a clear pricing structure. The image-to-video function has specific requirements for input photos, and users should be aware that the free tier is for personal, non-commercial use, with Sber retaining a broad license for generated content. AI
IMPACT Clarifies capabilities of Sber's GigaChat image tools, highlighting discrepancies between marketing and documentation for users.
RANK_REASON The article discusses the integration of new features into an existing product (GigaChat) and the confusion surrounding the underlying models, rather than a novel model release from a frontier lab.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →