Fine Tuning

1 min read
Training/1M tokensInference Input/1MInference Output/1M
GPT-4.1$25.00$3.00$12.00
GPT-4.1 Mini$5.00$0.80$3.20
GPT-4.1 Nano$1.50$0.20$0.80
o4-mini$100/hour$4.00$16.00
Consistent output format that prompting alone cannot achieve
Narrow domain expertise where few shot examples fall short
Reducing prompt length to lower per request costs
Teaching new facts (use retrieval augmented generation instead)
Replacing good prompt engineering for general tasks