[krackers]

I can believe this, Deepseek V3.2 shows that you can get close to "gpt-5" performance with a gpt-4 level base model just with sufficient post-training.

[irthomasthomas]

Deepseek scores Gold at IMO and IOI while GPT-5 scores Bronze. OpenAI now has to catch up to china.

[arcanemachiner]

...in a single benchmark.

[irthomasthomas]

No. Many benchmarks, I just mentioned those two as they where being bragged about by openai and Google when their internal models achieved gold.