Joe Fabisevich :verified: on Nostr: nprofile1q…gaynu That’s a fair enough point, but in the case of Llama it’s not ...

nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqwgx8ks7dhtw54n6tt0w6a0nk4fwh70wr0ltwzn43dqj9xh93h7cq9gaynu (nprofile…aynu) That’s a fair enough point, but in the case of Llama it’s not a reasoning model so that shouldn’t apply here. If it’s anything like GPT-4o, the cost should be in a similar ballpark of your chart.