Is GPT-4 Getting Faster? (Latency more than halved in the last 3 months)

API
, ,
1

Latency 50
Latency 505000×2500 672 KB

Seeing that GPT-4 latencies for both regular requests and computationally intensive requests have more than halved in the last 3 months.

Curious if others are seeing the same?

Wrote up my notes here.

1 Like