LLM API比較

ModelContext/OutputKnowledge Cutoff$Input/1M$Output/1M
gpt-4o-2024-05-13128k2023-10515
gpt-4-turbo-2024-04-09128k2023-121030
gpt-3.5-turbo-012516k2021-090.51.5
claude-3-opus-20240229200k/4k2023-081575
claude-3-sonnet-20240229200k/4k2023-08315
claude-3-haiku-20240307200k/4k2023-080.251.25
gemini-1.5-flash-latest1M/8kearly 20230 (0.35-0.7*)0 (0.53-1.05*)
gemini-1.5-pro-latest1M(2M)/8kearly 20230 (3.5-7*)0 (10.5-21*)
* >128k tokens