DeepSeek-V3 delivers answers with a median latency of just 0.8 seconds, compared to 1.6–1.8 seconds for GPT-4 Turbo — offering a 100% faster response time in practical deployments. This ...

As for what's needed to hit the next 10 times or 100 times jump in scale, Selsam, the OpenAI researcher, said it's data efficiency. The GPT models are very efficient at processing information ...