#ai-model-performance

[ follow ]
#gpt-54-thinking #chatgpt-capabilities #advanced-reasoning #text-analysis #gpt-52 #benchmarks #gdpval
Graphic design
fromZDNET
4 days ago

I tested GPT-5.4, and the answers were really good - just not always what I asked

GPT-5.4 Thinking delivers superior analytical depth and reasoning capabilities compared to earlier ChatGPT models, though formatting and image generation remain weaker areas.
Artificial intelligence
fromArs Technica
3 months ago

OpenAI releases GPT-5.2 after "code red" Google threat alert

GPT-5.2 improves performance and efficiency, outscoring competitors on multiple benchmarks and claiming much faster, dramatically cheaper completion of professional knowledge tasks than human experts.
[ Load more ]