ChatGPT’s new Images 2.0 model is surprisingly good at generating text
SMRTR summary
OpenAI's new ChatGPT Images 2.0 model can now accurately generate readable text within images — something AI image tools have historically struggled with. Unlike older diffusion models that often misspelled words, Images 2.0 uses "thinking capabilities" to double-check its work, support non-Latin languages, and produce detailed outputs like comic strips and marketing materials at up to 2K resolution. It becomes available to all ChatGPT users Tuesday.
SMRTR provides this summary for quick context. The original article belongs to TechCrunch.
Read the original article