GPT-image-1.5 model is now available
AI Impact Summary
GPT-image-1.5 is OpenAI's latest cutting-edge image generation model. It features improved performance, quality, editing controls, and face preservation. In editing mode, the model supports high input_fidelity and adding/removing one aspect of the input image while retaining others. Request access: limited access model application Key model capabilities: Includes all capabilities of GPT-image-1: Text to image generation Image to image generation (editing) Inpainting High quality image generations, up to 1024x1536 and 1536x1024 pixels Face preservation Follow the image generation how-to guide to get started with this model. Automatic speech recognition (ASR) model update gpt-4o-mini-transcribe-2025-12-15 Improved transcription accuracy and robustness for real-time scenarios. ~50% lower word error rate (WER) than previous gpt-4o-transcribe-mini on English benchmarks Improves multilingual performance across Japanese, Indic, and other languages. Reduced hallucinations on silence by up to 4×, making it a more reliable choice for noisy environments and real-world audio streams. Input remains audio, with text as output, and deployment is API-only. Realtime-mini (speech-to-speech) model update gpt-realtime-mini-2025-12-15 Feature parity with full gpt-realtime model in instruction-following and function-calling. Input and output are both audio, and is be API-only. Text to speech model update gpt-4o-mini-tts-2025-12-15 New benchmark for multilingual speech synthesis, More natural, human-like speech with fewer artifacts and improved speaker similarity. Input is text, output is audio, and deployment is API-only. October 2025 Realtime API support for SIP The Realtime API now supports SIP, enabling telephony connections to realtimeapi. For more information, see the Realtime SIP documentation . GPT-4o audio model released The gpt-4o-transcribe-diarize speech to text model is released. This is an Automatic Speech Recognition (ASR) model that converts spoken language into text in re