Open-Source

A high-quality text-to-image latent diffusion model trained on LAION-2B, enabling fast and flexible image generation.

A flagship text-to-image model with improved realism, composition, and support for high-resolution 1024x1024 image generation.

A multilingual speech recognition and translation model by OpenAI, supporting 100+ languages with improved robustness and low-latency transcription.