Daily Tech News: OpenAI Launches Images 2.5 and DeepSeek Cuts Flash API Prices
OpenAI announced ChatGPT Images 2.5 featuring up to 50% reduced generation latency alongside new sketch reference and in-image annotation tools. Additionally, DeepSeek announced an API price cut for its Flash AI model series, reducing output token costs to 4 RMB per million tokens. The release of ChatGPT Images 2.5 and DeepSeek's price cuts reflect intensifying competition across AI image generation and API inference efficiency. These updates lower the cost barrier for developers using LLM APIs while providing content creators with more precise image editing capabilities. ChatGPT Images 2.5 improves multi-turn editing consistency and prompt adherence, introducing features like template libraries and sketch input. DeepSeek's revised Flash API pricing takes effect on September 10, 2026, at 12:00 PM Beijing time.
## BACKGROUND
Conversational AI image generation tools often suffer from high latency and struggle to maintain subject consistency across sequential edit prompts. Meanwhile, providers of Mixture-of-Experts (MoE) 'Flash' models frequently lower API pricing to encourage developer adoption for high-volume agentic and coding workflows.