I’ve just launched Glazed.ai, an API-first platform (playground available) for creating AI characters that combine image generation and conversation, consistently and contextually.
Most tools treat image gen (Stable Diffusion) and LLM chat as separate experiences.
With Glazed, you define a character once, and interact through both chat and visuals - with stateful memory and consistent appearance.
What’s live:
•Image generation using Stable Diffusion 1.5 + community-favored models like EpicRealism
•Stateful conversations with memory per user and thread
•Lightweight API endpoints (/v1/images, /v1/conversations, etc.) — no polling or job queues
•Fast responses with smart defaults for image consistency and prompt hygiene
The product is fully usable now and actively being improved. Open to any feedback or questions.
- glazed.ai