Mastodon Feed
jsonstein@masto.deoan.org ("Jeff Sonstein") wrote:
Relative Activity & Testing Costs (gpt-4o vs gpt-4o-mini) October
gpt-4o
API Requests: 58
Tokens: 36,669
Dev Spend: $0.33 ($0.005689/Request)gpt-4o-mini
API Requests: 258
Tokens: 172,037
Dev Spend: $0.09 ($0.000346/Request)gpt-4o-mini used
~4.45 times the API Requests
&
~4.7 times the Tokens
but cost ~27% of using gpt-4oTLDR:
about 6/10ths of 1 cent per w gpt-4o
&
about 4/100ths of 1 cent per w gpt-4o-mini
&
no noticeable degradation in story quality