# ChangeLog ## v1.2.0 - 2026-10-22 ### Enhancements * Support GPT Image generation models, image model/quality/output-format setters, and streaming Base64 image fragments without buffering the full result in RAM. * Stream image responses with a host-tested incremental SSE parser and bounded heap workspace. ### Bug Fixes * Release the temporary HTTP payload after parsing image responses. * Use an explicit supported GPT Image resolution and retain legacy response-format defaults for compatible backends. ### Tests * Refresh the image-understanding fixture and audio model, and prevent main-task stack overflow in image-generation integration tests. * Run the live OpenAI API hardware test on ESP-IDF 5.5; retain compilation checks on 5.3, 5.4, and 5.5. ## v1.1.0 - 2025-7-24 ### Enhancements * Chat completion supports multi-modal input [#464](https://github.com/espressif/esp-iot-solution/pull/464) * Text-to-speech supports streaming output [#464](https://github.com/espressif/esp-iot-solution/pull/464) ## v1.0.2 - 2025-2-11 ### Bug Fixes: * Fix OpenAI_Request not handling chunked data correctly [#467](https://github.com/espressif/esp-iot-solution/pull/467) ## v1.0.1 - 2024-11-5 ### Bug Fixes: * Fix incorrect free of cjson object's payload memory ## v1.0.0 - 2024-5-15 ### Break Changes: * Changed C++ reserved keyword 'delete' to 'deleteResponse' ## v0.3.1 - 2023-12-29 ### Enhancements * Fix infinite loop of TTS read * Update Error message ## v0.3.0 - 2023-12-4 ### Enhancements * Support OpenAI text to speech api ## v0.2.0 - 2023-11-15 ### Enhancements * Support config the OpenAI Base URL through API `OpenAIChangeBaseURL` or `menuconfig`. Thanks [@Sam McLeod](https://github.com/sammcj) * Using MP3 format audio file in unit test ## v0.1.2 - 2023-8-4 ### Docs * Update OpenAI.h for build online docs ## v0.1.1 - 2023-6-29 ### Enhancements * Support http chunked transfer. ## v0.1.0 - 2023-6-14 ### Enhancements * Init version